U.S. patent application number 13/437833 was filed with the patent office on 2012-07-26 for incentive structure for centralized trading market.
This patent application is currently assigned to Automated Equity Finance Markets, Inc.. Invention is credited to Gregory Wayne DePetris, Martin Rein Hakker, Thomas Walter Little, JR., Sergei Poliakoff.
Application Number | 20120191592 13/437833 |
Document ID | / |
Family ID | 41465104 |
Filed Date | 2012-07-26 |
United States Patent
Application |
20120191592 |
Kind Code |
A1 |
DePetris; Gregory Wayne ; et
al. |
July 26, 2012 |
INCENTIVE STRUCTURE FOR CENTRALIZED TRADING MARKET
Abstract
An automated marketplace is separated into tiers, with
behavioral requirements for each tier. Tier eligibility is a
"structural incentive" for market participants to exhibit desirable
behavior and eschew undesirable behavior. Tiered eligibility also
reflects the natural imbalance of liquidity in the market and the
need to preserve the identity of a class of liquidity providers,
which further leads to a structure that can preserve the integrity
of person-to-person trading relationships even in an automated
environment. Within each tier, participant behavior leads to a
ranking for that participant. When specific events occur, these
events are allocated based on participant ranking. Participant
ranking is an "activity incentive" that influences the behavior of
market participants. Certain events are defined as desirable or
undesirable, and when performed by a market participant, lead to
positive incentives or negative incentives, of structural and/or
monetary type.
Inventors: |
DePetris; Gregory Wayne;
(Guilford, CT) ; Hakker; Martin Rein; (Norkport,
NY) ; Little, JR.; Thomas Walter; (Hoboken, NJ)
; Poliakoff; Sergei; (New York, NY) |
Assignee: |
Automated Equity Finance Markets,
Inc.
New York
NY
|
Family ID: |
41465104 |
Appl. No.: |
13/437833 |
Filed: |
April 2, 2012 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
12217456 |
Jul 2, 2008 |
8156023 |
|
|
13437833 |
|
|
|
|
Current U.S.
Class: |
705/37 |
Current CPC
Class: |
G06Q 40/00 20130101;
G06Q 30/02 20130101; G06Q 40/04 20130101; G06Q 30/0207 20130101;
G06Q 40/06 20130101 |
Class at
Publication: |
705/37 |
International
Class: |
G06Q 40/04 20120101
G06Q040/04 |
Claims
1. A method of allocating an event from an initiator to
counterparties using a stock loan trading system, comprising:
automatically determining, by a computer program executing at the
stock loan trading system, a ranked sequence of the counterparties,
based on activities of the counterparties during a first
predetermined time period, receiving, by the computer program
executing at the stock loan trading system, the event from the
initiator, allocating, by the computer program executing at the
stock loan trading system, the event according to the ranked
counterparties sequence, and moving, by the computer program
executing at the stock loan trading system, counterparties that
participated in the event to the end of the ranked counterparties
sequence.
2. The method of claim 1, wherein the event is a stock return, the
initiator is a borrower and the counterparties are lenders.
3. The method of claim 1, wherein the event is a stock recall, the
initiator is a lender and the counterparties are borrowers.
4. The method of claim 1, wherein the activities are one of loan
durations, loan values, and share quantity in the loans.
5. The method of claim 4, wherein each of the activities is
compared, by the computer program executing at the stock loan
trading system, with a low threshold, and wherein an activity that
is under the low threshold increases a frequency of the
corresponding one of the counterparties' representation in the
ranked counterparties sequence.
6. The method of claim 1, further comprising: determining whether
the initiator is eligible for a financial incentive, and when the
determination is positive, assigning, by the computer program
executing at the stock loan trading system, the financial incentive
to the initiator.
Description
[0001] This application is a divisional of U.S. patent application
Ser. No. 12/217,456 filed Jul. 2, 2008.
BACKGROUND OF THE INVENTION
[0002] The present invention relates to structural incentives,
activity incentives and economic incentives for participants in a
centralized trading market.
[0003] Externalities occur when one person's actions affect another
person's well-being and the relevant costs and benefits are not
reflected in market prices. Externalities can be positive,
providing benefits, or negative, providing harms. Automation of a
financial market has historically been difficult when the financial
instruments are highly illiquid or have naturally imbalanced supply
and demand. Thus, these markets remain dependent on
person-to-person trading, usually conducted via the telephone or
instant messaging. Market participants are resistant to losing the
benefits of person-to-person trading relationships. When
transitioning from a market premised on participants' knowledge of
each other, to an automated market where participants participate
anonymously, the behavior incentives due to personal relationships
are no longer available. Accordingly, to operate effectively,
automated incentives should take the place of the incentives
implicit in the personal relationships. In newly automated markets,
there is a need for automated incentives to promote good behavior
and discourage bad behavior.
SUMMARY OF THE INVENTION
[0004] In accordance with an aspect of this invention, there is
provided a method of allocating an event from an initiator to
counterparties using a stock loan trading system. A ranked
counterparties sequence is determined based on activities of the
counterparties during a first predetermined time period. An event
is received and is allocated according to the ranked counterparties
sequence, then counterparties that participated in the event are
moved to the end of the ranked counterparties sequence.
[0005] In one version of this invention, the event is a stock
return, the initiator is a borrower and the counterparties are
lenders.
[0006] In another version of this invention, the event is a stock
recall, the initiator is a lender and the counterparties are
borrowers.
[0007] The activities may include loan durations, loan values, and
share quantity in the loans.
[0008] The computer program executing at the stock loan trading
system may compare each of the activities with a low threshold, so
that an activity that is under the low threshold increases a
frequency of the corresponding one of the counterparties'
representation in the ranked counterparties sequence.
[0009] The computer program executing at the stock loan trading
system may assign a financial incentive to the initiator when the
initiator is determined to be eligible for a financial
incentive.
[0010] It is not intended that the invention be summarized here in
its entirety. Rather, further features, aspects and advantages of
the invention are set forth in or are apparent from the following
description and drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
[0011] FIG. 1 is a block diagram showing the conventional
relationships between a borrower, a lender and a prime broker for a
securities loan;
[0012] FIG. 2 is a block diagram showing entities conventionally
participating in a securities loan and a securities trade;
[0013] FIG. 3A-3B are a flowchart showing conventional actions
involved in a securities loan for a short sale and the associated
short sale trade;
[0014] FIG. 4 is a block diagram showing a borrower, a lender and
an automated market for securities loan;
[0015] FIG. 5 is a more detailed block diagram of the entities in
FIG. 4;
[0016] FIG. 6 is a chart showing hypothetical loan profiles for
four different market participants;
[0017] FIG. 7 is a chart showing daily activity for four
hypothetical market participants;
[0018] FIGS. 8A-8C are a flowchart showing automated establishment
of a securities loan;
[0019] FIGS. 9A-9C are a flowchart showing automated incentives for
a borrower; and
[0020] FIGS. 10A-10B are a flowchart showing automated incentives
for a lender.
DETAILED DESCRIPTION
[0021] Finding ways to affect trading programs in similar manner as
relationships affect person-to-person trading is important as
demands for "transparency" increase which in turn place scrutiny on
person-to-person trading.
[0022] Desirable behavior refers to activity that leads to a
marketplace with desirable characteristics such as deeper
liquidity, smaller buy-sell quote spreads, reduced systemic risk
and more equitable access for market participants. Undesirable
behavior leads away from a marketplace with desirable
characteristics. Undesirable behavior also includes using the
marketplace only for price discovery ("gaming" or "bypass") with
intent to execute in another marketplace. In particular, gaming
behavior is characterized by chronic sending of orders to the
marketplace, determining as much price information as possible,
then cancelling the orders generally to the detriment of other
market participants.
[0023] An automated marketplace is separated into tiers, with
behavioral requirements for each tier. Tier eligibility is a
"structural incentive" for market participants to exhibit desirable
behavior and eschew undesirable behavior. Tiered eligibility also
reflects the natural imbalance of liquidity in the market and the
need to preserve the identity of a class of liquidity providers,
which further leads to a structure that can preserve the integrity
of person-to-person trading relationships even in an automated
environment.
[0024] Within each tier, participant behavior leads to a ranking
for that participant. When specific events occur, these events are
allocated based on participant ranking Participant ranking is an
"activity incentive" for market participants to exhibit good
behavior and eschew bad behavior.
[0025] Certain events are defined as desirable or undesirable, and
when performed by a market participant, lead to positive incentives
or negative incentives, of structural and/or monetary type.
Incentives encourage market participants to exhibit good behavior
and eschew bad behavior.
[0026] Automated structural, activity and economic incentives will
now be discussed in the context of the securities lending
marketplace. Use of incentives is not limited to securities
lending. Incentives are useful in a wide variety of situations such
as pollution rights trading, derivative financial markets that
provide liquidity rebates (ex: ISE, ARCA) and so on.
[0027] Incentives are not limited to the disclosed types of
incentives.
[0028] The conventional U.S. securities lending marketplace will
now be discussed.
[0029] A normal or "long" sale is the sale of a security that the
seller presently owns.
[0030] A "short" sale is the sale of a security that the seller
does not own or any sale that is consummated by the delivery of a
security borrowed by, or for the account of, the seller. Usually, a
short seller expects the market price for a security to decrease; a
short seller sells now, expecting to buy at a lower price in the
future to close out her position. This is a profitable strategy
when it achieves the sequence of sell high then buy low. The ease
of short selling is crucial for effective arbitrage. Short sellers
generally do not know how long they will maintain their
position.
[0031] Short sellers include hedge funds, mutual funds (if
permitted by the rules of the fund), institutional investors,
retail investors, brokers trading for their own account,
arbitrageurs, market makers, risk managers, speculators, and so
on.
[0032] The major reason that someone wants to borrow securities is
to accomplish a short sale in compliance with the Securities and
Exchange Commission (SEC) regulations. Specifically, SEC Regulation
SHO requires short sellers in all equity securities to locate
securities to borrow before selling, see
http://www.sec.gov/rules/final/34-50103.htm. Other reasons for
borrowing securities, referred to as "permitted purposes" under
Regulation T of the Board of Governors of the Federal Reserve
System, include (i) to prevent a settlement failure, and (ii) for
establishing an Exchange Traded Fund (ETF). The permitted purpose
need not be accomplished by one of the parties, but must occur
somewhere in an associated transaction. Firms with a large number
of active retail accounts and substantial revenue are exempt from
the permitted purpose regulation and so can borrow for any
reason.
[0033] Securities lenders, or their agents, are parties that
presently own the security. The legal owner of a security is
referred to as the "beneficial owner". Lenders wish to lend to make
profit on their securities inventories that are otherwise idle. Big
custody banks are the largest lenders in the US market, lending as
agents on behalf of large institutional owners such as pension
funds, public retirement funds, mutual funds and endowments.
Additionally, brokers want to lend to enable their customers to
accomplish short sales. Securities lending and margin finance are
responsible for over half of prime brokerage revenues. At end 2007,
US$2.1 trillion of equities were on loan in global markets.
[0034] Some security owners find short selling distasteful and will
not lend as they believe short selling facilitates downward price
pressure, thereby devaluing their inventory.
[0035] If not for certain "Prohibited Transaction Exemptions"
issued by the U.S. Department of Labor, employee benefit plans
would refrain from lending to avoid violating provisions of the
Employee Retirement Income Security Act of 1974 ("ERISA").
[0036] A short sale consists of a trader selling stock that the
trader does not own on trade day (T), and delivering borrowed stock
on settlement day, which is the third day after the trade day
(T+3). Eventually, the trader closes her position by buying stock,
and terminating the stock loan. That is, the trade occurs on T,
while the stock loan occurs on T+3. The trade settlement also
occurs on T+3.
[0037] To bridge the time difference between T and T+3, "locate"
practice is used in the securities industry. A locate is an
affirmative determination that a party will provide the named
quantity of securities three days hence.
[0038] A trader can obtain a locate by asking a broker, such as by
telephone, email or instant message, or by consulting a locate file
provided by the broker to the trader each morning, listing the
inventory that the broker has available to loan. Not all locates
are actually converted into loans. Reasons for non-conversion
include that the stock is either not needed as the short seller
closed her position prior to T+3 or that the stock was actually
loaned by a party other than the locate provider.
[0039] A broker maintains lendable securities inventories when the
broker trades for its own account, and when the broker holds
securities on behalf of the margin accounts of customers who have
bought the securities. Section 8 of the Exchange Act of 1934
prohibits brokers from lending shares held in retail cash accounts
or retail non-margin accounts. Usually, when a retail customer
opens a brokerage account, the terms of the account permit the
broker to re-hypothecate and lend securities that the customer
holds.
[0040] In exchange for a loan, the customer provides cash
collateral in an amount slightly greater than the value of the
securities, such as $102 of cash for each $100 of securities value.
The broker pays interest to the securities borrower on the cash
collateral. The "rebate rate" is the interest rate paid for the
cash collateral. Negative rebates can and do occur, corresponding
to expensive loans.
[0041] A securities loan is for a period of one-day and is
self-renewing ("overnight self-renewing) unless either (a) the
borrower returns the shares of the security, or (b) the lender
recalls the shares of the security.
[0042] If the value of the security fluctuates, the amount of
required cash collateral correspondingly fluctuates, so the
borrower may have to supply more money or may receive money
back.
[0043] If the broker lacks its own inventory to make a securities
loan, the broker finds a lender, then the broker enters into a
first securities loan contract with the lender, and then enters
into a second securities loan contract with the borrower. The terms
of the loan contracts are different, theoretically providing profit
to the broker for its services in arranging the loan.
[0044] Generally, the loans have standardized terms and conditions,
but the interest rates are different depending on the relationships
between the parties (long-standing relationship vs. first
transaction), the amount of stock being loaned (small loans tend to
be more expensive per share) and the characteristics of the
security (readily available vs. hard to find), and so on.
[0045] In the U.S., the Securities Industry and Financial Markets
Association (SIFMA) provides and updates standardized terms for
securities loans as the Master Securities Loan Agreement (MSLA),
available at
www.sifma.org/services/stdforms/pdf/master_securities_loan_agreement.sub.-
--2000_version.pdf
[0046] In Europe, the International Securities Lending Association
(ISLA) provides an updates standardized terms for securities loans
as the Global Master Securities Loan Agreement (GMSLA), available
at http://www.isla.co.uk/docs/Gmsla %202000%20version.doc.
[0047] FIG. 1 is a block diagram showing the conventional
relationships between borrower 10, prime broker 20 and lender 30
for a pair of securities loans.
[0048] The first loan involves, at action A, borrower 10 providing
cash collateral to broker 20, such as $105 per $100 of securities
value. At action B, broker 20 provides a loan of the security
shares to borrower 10, and at action C, broker 20 provides interest
on the cash collateral to borrower 10. The interest is computed and
credited to borrower 10 on a daily basis. The interest is expressed
relative to the Federal Funds (FF) overnight interest rate, i.e.,
FF minus bb basis points.
[0049] The second loan involves, at action D, broker 20 providing
cash collateral to lender 30, such as $102 per $100 of securities
value. At action E, lender 30 provides a loan of the security
shares to broker 20, and at action F, lender 30 provides interest
on the cash collateral to broker 20. The interest computed and
credited to borrower 10 on a daily basis. The interest is expressed
relative to the Federal Funds overnight interest rate, i.e., FF
minus ee basis points.
[0050] Broker 20 can make profit (or loss) from the difference in
cash collateral between the first and second loans, and from the
difference in interest paid on the cash collateral between the
first and second loans.
[0051] The price of a stock loan means the interest rate paid to
the borrower on the cash collateral posted to the lender. The
borrower is not actually paying anything, but rather, choosing to
accept more or less interest on the collateral.
[0052] From the borrower's perspective, "to pay more" means to
accept a low interest rate on the loan. For easy to borrow, widely
available securities, borrowers expect to be paid higher interest
rates on their cash collateral. The borrower has many choices for
sources of the stock, and is thus unwilling to pay a premium for
the inventory.
[0053] From the lender's perspective, "to pay more" means to pay
higher interest on the collateral. Lenders expect to pay higher
interest rates when loaning very liquid securities. Ultimately, the
lender profits are the difference between the cash reinvestment
rate they earn internally relative to the rate paid to the
borrower, or the difference between the interest rate the lender
borrowed the stock at, and the rate they lend it at. In either
case, a profit-maximizing lender chooses to offer securities at the
lowest market clearing interest rate.
[0054] FIG. 2 is a block diagram showing entities conventionally
participating in a securities loan and a securities trade. On trade
day T, borrower 10 obtains a locate to engage in a short sale of
stock shares. Prime broker 20 provides the locate to borrower 10.
If necessary, prime broker 20 obtains inventory for the locate via
a second locate with lender 30. Borrower 10 sends its short sale
order to executing broker 60, which relays it to exchange 70.
Exchange 70 matches the short sale order with a buy order from
buyer 90, relayed to exchange 70 via executing broker 80.
[0055] Prime broker 20, executing broker 60 and executing broker 80
are shown as different entities. In practice, one firm may fulfill
one, two or three of these roles.
[0056] On settlement day, T+3, executing broker 60 actually makes
the loan to borrower 10, from its inventory or if necessary, by
borrowing stock from lender 30 (the same or a different instance of
lender 30 that may have participated in the locate).
[0057] FIGS. 3A and 3B are a flowchart showing conventional actions
involved in a securities loan for a short sale and the associated
short sale trade, as generally described above. FIG. 3A shows
activity on trade day T. FIG. 3B shows activity on settlement day
T+3.
[0058] At step 100, borrower 10 requests a locate from broker 20,
such as by calling broker 20 or sending an e-mail to broker 20. In
some embodiments, borrower 10 checks a locate file supplied to her
each morning by broker 20. Borrower 10 can manually check the
locate file, or the execution management system (EMS) being used by
borrower 10 to enter a short order can automatically check the
locate file and append the Locate ID to the short order.
[0059] Meanwhile, at step 101, buyer 90 sends a buy order for the
security that borrower 10 is interested in to executing broker 80.
At step 102, broker 80 receives the buy order and relays it to
exchange 70. At step 103, exchange 70 receives the buy order.
[0060] At step 105, broker 20 receives the locate request. At step
110, broker 20 checks its stock inventory. The checking may occur
via a person consulting an inventory database, or by a computer
system checking a locate file. If broker 20 has sufficient
inventory to provide the requested locate, then action continues at
step 135. If broker 20 lacks sufficient inventory to provide the
requested locate, then at step 115, broker 20 requests a locate,
either by calling potential lenders on the telephone, by emailing
potential lenders, or by checking an online system with locate
files from third parties.
[0061] At step 120, lender 30 receives the locate request from
broker 20, and at step 125, provides the locate. At step 130,
broker 20 receives the locate from lender 30.
[0062] At step 135, broker 20 sends a locate to borrower 10. The
locate identifies broker 20 and enables broker 20 to locate the
specific shares promised for the loan to borrower 10. At step 140,
borrower 10 receives the locate ID from broker 20. At step 145,
borrower 10 sends a short sale order to executing broker 60,
including the locate ID. At step 147, broker 60 relays the short
sale order to exchange 70.
[0063] At step 150, exchange 70 receives the short sale order. At
step 155, exchange 70 matches the short sale order received at step
150 with the buy order received at step 130. At step 160, exchange
70 sends an execution report to each of executing broker 60 and
executing broker 80.
[0064] At step 162, executing broker 80 relays the execution report
to buyer 90.
[0065] At step 163, executing broker 60 relays the execution report
to borrower 10.
[0066] At step 166, borrower 10 receives the execution report from
executing broker 60.
[0067] Turning to FIG. 3B, on day T+3, at step 170, the computer
system for executing broker 60 realizes that a loan is needed to
enable settlement of the short sale executed on day T. At step 175,
the computer system for broker 60 checks its inventory. If broker
60 has sufficient inventory to make the loan, processing continues
at step 199.
[0068] If broker 60 lacks sufficient inventory to make the loan,
then at step 180, broker 60 requests a stock loan from lender 30.
At step 185, lender 30 receives the stock loan request and at step
190, provides the securities loan and decrements its inventory of
lendable stock. At step 195, broker 60 receives the securities
loan.
[0069] At step 198, broker 60 makes a securities loan to borrower
10 and decrements its inventory of lendable stock. At step 199,
borrower 10 receives the securities loan.
[0070] As mentioned, the securities loan is usually an overnight
self-renewing loan. On day T+4, interest starts to be paid on the
cash collateral from the collateral holder to the collateral
provider. The daily interest rate is usually expressed relative to
the Federal Funds rate which can change daily.
[0071] A small amount of the securities loan market occurs in a
centralized clearinghouse environment called the OCC Stock Loan
Program, a trade reporting facility that allows OCC's clearing
members to use borrowed and loaned securities to reduce OCC margin
requirements. In this program, the loan is legally between the
borrower and lender, that is, non-anonymous, and OCC guarantees
mark-to-market payment between the program participants.
[0072] So-called "processing systems" exist, such as Equilend and
Loanet. Processing systems consolidate and track information about
stock loans, typically negotiated by telephone and then entered to
a processing system to avoid keeping paper activity records.
Processing systems also perform other functions.
[0073] The above-described securities loan market is evolving as
new types of clients have come into existence, is fraught with
inefficiency and has problems.
[0074] A situation in the securities loan market is that as the
years go by, margins in the securities loan business have been
getting thinner because the financing costs paid by borrowers has
been generally decreasing while infrastructure costs have remained
approximately constant; meanwhile, the business is capital
intensive and requires specialized staff who understand the arcane
practices in the business.
[0075] Another situation in the securities loan market is the
emergence of high frequency traders such as day traders and
statistical arbitrageurs.
[0076] So called "day traders" close out their positions at the end
of each day. A day trader may execute a short sale in the morning,
and will then buy the stock before the end of the day. In these
situations, a loan will never occur; nevertheless, a locate is
required for the short sale. Day traders bring the overhead of a
locate but no chance of a loan. An automated service,
www.locatestock.com, fills this niche, that is, brokers that grant
locates via this automated service charge a fee for each locate,
expecting that a loan will not occur.
[0077] So called "statistical arbitrageurs" use statistical
techniques to exploit trading opportunities that are usually
intra-day, but can be longer.
[0078] In the remainder of this document, the needs of non-settling
borrowers are ignored.
[0079] Another problem in the securities loan market is that some
customers are using trading strategies that cause them to close
their short positions very quickly, such as within a few days. If
borrower 10 returns the stock after only a few days, say on T+4
through T+8, then the lender(s) have virtually no time to make
profit, that is, the slim profit they make just about covers their
expenses. Reasons for stock returns include closing some or all of
the short position, and finding a cheaper stock loan.
[0080] From the viewpoint of a lender, stock returns are
undesirable behavior. In contrast, borrowers who borrow for a long
time are desirable customers.
[0081] A substantial problem for a short seller is a lender recall
of the securities loan. As permitted in the standard securities
lending agreement, lender can take back, or recall, its stock for a
variety of reasons: to sell the stock according to a trade
decision, to lend the stock to a different customer, to participate
in a shareholder vote, or because the lender hasmust obtain
"possession and control" of customers' fully-paid and excess margin
securities out on loan to comply with SEC Rule 15c3-3 (customer
protection) and/or SEC Rule 15c3-1 (capital requirements for a
firm). Recall rights sometimes exist for tax purposes: the IRS
(Section 1058) requires a recall provision for manufactured
dividend payments to remain nontaxable income (for certain exempt
funds) and for the loan not to be treated as a sale.
[0082] Usually, broker 20 tries very hard to find substitute stock
for borrower 10. If substitute stock cannot be found, then borrower
10 is forced to close ("cover") its short position immediately,
which may wreak havoc with its trading strategy, and lead to a big
loss during a "short squeeze", that is, a situation in which the
price of the stock rises and investors who sold short rush to buy
it to cover their short position and cut their losses. As the price
of the stock increases, more short sellers feel compelled to cover
their positions. In many cases, borrower 10 cannot simply create a
new short position to replace the closed position as a new loan is
unavailable. Borrowers consider a forced closing of their short
position during a falling market to be a horrible event.
[0083] From the viewpoint of a borrower, stock recalls are
disruptive events. Borrowers assert that prime brokers who protect
the borrowers from recalls, by finding substitute stock for
recalled stock, provide a valuable service. In contrast, lenders
who lend as long as the borrower desires are trustworthy and
preferred. The largest prime brokers state that they operate for
years without recalling securities loans to their clients.
[0084] Generally, borrowers face the following challenges in the
securities loan market: first, finding the stock; second, whether
the loan is stable, i.e., not subject to a recall; third, getting
the loan at as low a price as possible subject to stability; and
fourth, whether the counterparty is creditworthy.
[0085] Thus, there is room for improvement in the securities loan
market.
[0086] An automated marketplace for securities lending will now be
discussed.
[0087] FIG. 4 is a block diagram showing borrower 15, lender 35 and
electronic loan market system (ELMS) 200. ELMS 200 is a computer
system having one or more general purpose computers executing
software for performing its functions, as discussed below, along
with suitable communication facilities for its users, also referred
to a market participants, specifically lender 35 and borrower 15,
and suitable memory and storage.
[0088] Generally, lender 35 makes stock inventory available to ELMS
200. Lender 35 may be the parties discussed above as suitable for
lender 30 or broker 20. ELMS 200 maintains a record of available
inventory, by security. Borrower 15 sends a loan request to ELMS
200 and, after competing for the loan in ELMS 200, receives a loan
commitment and a locate ID. Borrower 15 may be the parties
discussed above as suitable for borrower 10 or broker 20.
[0089] The loan involves, at action X, borrower 15 providing cash
collateral to lender 35, such as $103 per $100 of securities value.
At action Y, lender 35 provides a loan of the security shares to
borrower 15, and at action Z, lender 35 provides interest on the
cash collateral to borrower 15. The interest payable is calculated
daily, accrued, and paid at the end of the month to borrower 15.
The interest is expressed relative to the Federal Funds (FF)
overnight interest rate, i.e., FF minus zz basis points.
[0090] The identities of the parties to a loan are not known to
each other. Accordingly, the personal relationships in the
conventional securities loan market that deter undesirable behavior
are entirely absent in the environment of FIG. 4, resulting in
severe risk of negative externalities.
[0091] Automated structural, activity and economic incentives are
provided in ELMS 200 to deter bad behavior and promote good
behavior. Bad behavior refers to lender stock recalls and borrower
stock returns, and other "gaming" behavior such as high frequency
of cancellation of offers to borrow or lend, overly aggressive
re-rates (a "re-rate" is a request to change the rate of a loan,
initiated by either the borrower or seller, and sent to the
universe of participants involved in lending the stock). Good
behavior refers to lenders not recalling stock, and borrowers
keeping the loan outstanding for long durations and having a low
frequency of order cancellation and eschewing re-rates.
[0092] Structural incentives will now be discussed.
[0093] As a structural incentive, ELMS 200 is separated into two
tiers, primary market 210 and secondary market 220. In other
embodiments, three or more tiers may be provided. Primary market
210 is intended for participants exhibiting good behavior, that is,
to replicate the stability available in the conventional
un-centralized (distributed) securities loan market. Secondary
market 220 is intended for all other participants, i.e., those
exhibiting generally reasonable behavior. Secondary market 220 is
suitable for smaller lenders who are comfortable operating
differently than conventional large lenders.
[0094] The costs of obtaining a securities loan in primary market
210 are more than the costs in secondary market 220 to compensate
lenders for expected stability. In other words, some borrowers
prefer to pay a premium for stable supply and choose the more
stable primary market 210 relative to the less stable secondary
market 220. Thus, the primary market is structured to promote
desirable behavior.
[0095] The process is identical for market tiers, with the
differences being (i) the expectations of stability and price, and
(ii) the eligible participant list for each market. The primary and
secondary markets operate independently with separate auctions.
[0096] Inventory can be transferred between the primary and
secondary pools depending on the access level of participants in
the system. For example, it is possible for a borrower to remove
inventory from the primary market to lend it to the secondary
market, if the economics are favorable.
[0097] FIG. 5 is a more detailed block diagram of the entities in
FIG. 4. ELMS 200 comprises primary market 210, secondary market
220, and associated facilities for record-keeping and reporting.
ELMS 200 is a general purpose computer or computers that cooperate
to execute a software program or programs according to the present
invention. The entities in FIG. 5 communicate via wireline or
wireless communications, using suitable ones of dedicated
communication channels, private networks and/or public networks.
ELMS 200 is provided with suitable equipment, such as memory,
storage (e.g., magnetic, optical, magneto-optical or other suitable
storage), input peripherals (e.g., keyboard, voice input,
communication channel input) and output peripherals (e.g.,
displays, printers). ELMS 200 is provided with suitable software
infrastructure, such as operating system, communication channel
drivers, device drivers and so on.
[0098] Securities loans arranged through ELMS 200 are automatically
reported by ELMS 200 to trade reporting facility 40 and clearing
entity 50. ELMS 200 also provides a facility (not shown) for its
participants to report manually negotiated securities loans to
trade reporting facility 40 and clearing entity 50. Trade reporting
facility 40 may be an existing processing service such as Equilend
or Loanet. Clearing entity 50 may be one or more of Options
Clearing Corporation (OCC), Depository Trust Clearing Corporation
(DTCC), National Securities Clearing Corporation (NSCC), Boston
Stock Exchange Clearing Corporation, Philadelphia Stock Exchange
Clearing Corporation, or other suitable SEC regulated CA-1 facility
that can clear security trades.
[0099] ELMS 200 arranges and records loans.
[0100] Clearing entity 50 is the counterparty to each loan.
[0101] Clearing entity 50 obviates the conventional privity between
borrower and lender. Privity is a direct relationship between
parties to a contract or transaction sufficient to support a legal
claim. Benefits include: (i) operationally simpler--no separate
loan agreement for each loan; (ii) more cost effective--reduced
legal costs; and (iii) anonymous. Because ELMS 200 is involved in
arranging each transaction, it can allocate activity according to
an incentive system, and can be structured to facilitate
incentives.
[0102] A lender is classified as one of primary liquidity provider
(PLP) 230, competitive liquidity provider (CLP) 240 and electronic
participant (EP) 250. Initially, a lender is assigned to one of
these three categories, and over time, if the lender does not
conform to the behavior required for the category, the lender may
be re-assigned to another category. As discussed below, the lenders
in each category are also ranked within the category. In other
embodiments, different categories may be provided, such as a
further category PLP+.
[0103] A borrower is one of CLP 241 and EP 251. By definition of a
PLP, a PLP is only a lender. Examples of a PLP include insurance
companies and pension funds. An EP is permitted to lend only in the
secondary market. CLP 240 and CLP 241 are entities in the same
category, but one is acting as a lender and the other as a
borrower. A CLP entity can be either a lender or a borrower over
the course of its life, but in a particular transaction it assumes
one role. EP 250 and EP 251 are, similarly, entities in the same
category but one is acting as a lender and the other as a
borrower.
[0104] Borrowers generally request a loan to either (i) refinance
an existing (already settled) short position, or (ii) provide
inventory for an executed short order that is settling.
[0105] The specific category characteristics for a PLP, CLP and EP
are outside the scope of the present application.
[0106] Another structural incentive is the excellent credit rating
of clearing entity 50 that guarantees the daily mark-to-market of
cash flows in the event of counter-party default. For example,
assume that a borrower provided $102 cash for $100 of securities,
and that the securities lender then went out of business and did
not return the cash. Without a guarantee, the borrower would lose
$2 plus any increase in the market value of the securities. The
borrower loses the difference between the cash posted and the stock
price, therefore if the stock price goes down the borrower loses
more money as the collateral he holds is worth less. Of course, if
the securities had increased in value sufficiently, the borrower
would have a net gain. With a guarantee, clearing entity 50
reimburses the borrower for her loss, if any.
[0107] Activity incentives will now be discussed.
[0108] U.S. Pat. No. 6,618,707 (Katz) discloses a system for
automating options trading in which an incoming order is filled
against quotations based on the size of the quotations, as an
incentive for members to provide quotations of more than the
minimum size. This is an example of a positive activity
incentive.
[0109] Another example of a positive activity incentive is the
practice of certain prior art marketplaces paying participants for
order flow.
[0110] Within each market tier of ELMS 200, participant behavior
leads to a ranking for that participant. When specific events, such
as new loans, returns, recalls, re-rates and so on occur, these
events are allocated based on participant ranking. Desire to avoid
unwanted events, and to receive desired events, leads participants
to care about their ranking. This is similar to how personal
relationships induce people to behave better. Studies of eBay's
feedback system indicate that the mere existence of a mechanism
that monitors behavior can improve performance of parties using a
transactional system.
[0111] The ranking may be a unique sequential rank within a
category of participant or marketplace tier, or may be a
market-wide (global) level of rank, such as "superior", "normal",
"poor" and so on.
[0112] An example of how a participant's behavior can change their
ranking is now discussed.
[0113] Assume that rank is a unique sequential number, and category
is PLP, and there are five PLP participants: PLP-1, PLP-2, PLP-3,
PLP-4 and PLP-5. Further assume that, at the start of the day,
their ranking was as shown in Table 1, corresponding to the number
of loan recalls ever initiated by the participant.
TABLE-US-00001 TABLE 1 Rank Participant No. recalls 1 PLP-1 6 2
PLP-2 4 3 PLP-3 2 4 PLP-4 1 5 PLP-5 0
Let it be assumed that during the day, there were only two recalls
in primary market 210 and both recalls were from PLP-4 that
formerly had only one (1) recall. At the end of the day, ELMS 200
adjusts the rankings so that PLP-4, with three (3) recalls, has a
higher rank as shown in Table 2.
TABLE-US-00002 TABLE 2 Rank Participant No. recalls 1 PLP-1 6 2
PLP-2 4 3 PLP-4 3 4 PLP-3 2 5 PLP-5 0
In other embodiments, the recalls are measured relative to a moving
window, for example, the last two weeks, or percentage of the last
100 loans, or any other suitable metric.
[0114] The rankings operate as a sort of automated Golden Rule: do
unto ELMS 200 as you would have it do unto you. That is, the more
undesirable behavior (loan recalls) that a PLP participant
initiates, the more undesirable behavior (stock returns) will the
PLP be subject to.
[0115] Economic incentives will now be discussed.
[0116] Certain events are defined as desirable or undesirable, and
when performed by a market participant, incur monetary incentives
or disincentives. Here, it is useful to define behavior as relative
to a benchmark for what is normal for a category of market
participant.
[0117] FIG. 6 is a chart showing hypothetical loan duration
profiles for four different market participants where the term of
the loan was ended by the participant. The abscissa (X-axis) is
loan duration in days; the ordinate (Y-axis) is how many loans have
the specified duration. Curve AA shows a high frequency trader,
such as a hedge fund executing automated programs resulting in
frequent buys and sells. As shown, curve AA has an average loan
duration of 5 days. Curve BB shows a broker lending stock from its
own inventory (from its own account or held on behalf of retail
customers), the loans having an average duration of 15 days. Curve
CC shows a so-called long-short trader, such as a hedge fund,
having an average loan duration of 30 days. Curve DD shows a
pension fund that can readily accommodate long duration loans,
shown as having an average of 60 days.
[0118] Generally, the parties represented by curves BB and DD
should use primary market 210, while the parties represented by
curves AA and CC should use secondary market 220.
[0119] For curve DD, bad behavior is represented by short duration
loans, the leftmost tail of the curve. The cutoff is set as, for
example, the number of days that is two standard deviations from
the average length (.mu.-2.sigma.), or the number of days such that
5% of the loans are shorter than that number, or any other suitable
threshold.
[0120] For curve DD, good behavior is represented by long duration
loans, the rightmost tail of the curve. The cutoff is set as, for
example, the number of days that is two standard deviations from
the average length (.mu.-2.sigma.), or the number of days such that
5% of the loans are longer than that number, or any other suitable
threshold.
[0121] Good and bad behavior are defined similarly for each of
curves AA, BB and CC.
[0122] Although FIG. 6 assesses the number of loans, in other
embodiments, instead, the value of the loans is assessed, or the
number of shares loan. Generally, the profiles are computed for
each security over a moving window of time, with the window varying
by security, that is, thinly traded stocks have a longer window
such as one month, while actively traded stocks have a shorter
window such as one week.
[0123] FIG. 7 is a chart showing daily activity for four
hypothetical market participants. Curve AAA shows that its market
participant terminated, for instance, 7 loans after 4 days, 9 loans
after 5 days and 11 loans after 5 days. Comparing curve AAA in FIG.
7 with curve AA in FIG. 6, it is seen that the activity represented
by curve AAA is "normal" for curve AA, and so market participant
AAA will not get incentives.
[0124] Comparing curve BBB in FIG. 7 with curve BB in FIG. 6, it is
seen that the activity represented by curve BBB is "normal" for
curve BB, and so market participant BBB will not get
incentives.
[0125] Comparing curve CCC in FIG. 7 with curve CC in FIG. 6, it is
seen that the activity represented by curve CCC shows an
undesirably short loan of duration about 23 days, and some
desirably long loans of duration about 35 and 36 days. Market
participant CCC should get a negative incentive and two positive
incentives.
[0126] Comparing curve DDD in FIG. 7 with curve DD in FIG. 6, it is
seen that the activity represented by curve DDD shows some
undesirably short loans of duration under 50 days, and some
desirably long loans of duration over 70 days. Market participant
DDD should get negative incentives and positive incentives.
[0127] Operation of the incentives in ELMS 200 will now be
discussed in detail.
[0128] FIGS. 8A-8C are a flowchart showing automated establishment
of a securities loan.
[0129] At step 300, PLP 230 submits its stock inventory available
for lending to ELMS 200. At step 305, ELMS 305 receives the stock
inventory and stores it in a data file.
[0130] At step 310, EP 251 request a stock loan from ELMS 200. At
step 315, ELMS 200 receives the loan request, and checks its data
files for available inventory. In this case, ELMS 200 finds the
inventory from PLP 230 that is in the correct stock and of
sufficient quantity to support the requested loan, and determines
that no other lenders have suitable inventory.
[0131] At step 320, ELMS 200 allocates the inventory to EP 251,
discussed in detail with regard to FIG. 8C.
[0132] At step 322, ELMS 200 decides whether to assign positive or
negative financial incentives to the borrowers and lenders
participating in the stock loan match. Generally, if incentives are
earned relative to the auction match, the incentives are positive,
as it is desirable to encourage match activity. An example of a
positive financial incentive is a rebate on marketplace fees
imposed by ELMS 200.
[0133] At step 325, ELMS 200 sends a loan match report for the
allocated inventory to EP 251, trade reporting facility 40 and
clearing corporation 50. At step 330, EP 251 receives the loan
match report for its requested loan. At step 335, clearing
corporation 50 receives the loan match report. At step 340, trade
reporting facility 40 receives the loan match report.
[0134] In other cases, inventory from multiple lenders can be used
to fulfill the stock loan request.
[0135] In other embodiments, instead of lenders, such as PLP 230,
providing an inventory list at the start of the day, the lenders
register as wanting to be advised when there is a loan request
within a specified quantity range for various symbols. ELMS 200
then broadcasts the loan request to registered lenders. ELMS 200
selects interested lenders according to a procedure, such as
waiting a predetermined time for indications from lenders, then
selecting the lender of best rank at the best price.
[0136] At the end of the day, at step 350, ELMS 200 collects
activity information for all market participants and updates their
respective rankings with their category, discussed in detail with
regard to FIG. 8B. ELMS 200 also flags participants whose category
should be altered, to human analysts. In this embodiment, a
lender's behavior can indicate it should be changed from CLP status
to PLP status, or from PLP status to CLP status; in other
embodiments, different alterations occur. Or, the behavior could
result in access privileges being changed from primary to secondary
access.
[0137] FIG. 8B provides detail for step 350 of FIG. 8A.
[0138] At step 352, market participants, i.e., borrowers and
lenders, are rated based on recent activity, as discussed above.
The outcome is a ranking or rating for each participant, such as
"good", "normal" or "poor".
[0139] At step 353, ELMS 200 compares the participant's behavior
relative to "tier-normal" behavior. As used herein and in the
claims, "tier-normal" refers to behavior that is appropriate for
the tier, based on the actual behavior of other participants in the
tier and/or a hypothetical profile for the tier. FIG. 6 shows
hypothetical profiles for different participants in different
tiers. When a behavior is within the positive and negative
thresholds, it is normal for the tier.
[0140] At step 354, ELMS 200 produces a report suggesting which
participants, if any, should be changed to a different market tier.
Participants below the negative threshold for tier-normal behavior
are candidates for a lower tier. Participants above the positive
threshold for tier-normal behavior are candidates for a higher
tier. In the present embodiment, this decision is made by a human;
in other embodiments, the decision is made by ELMS 200, using a
decision criterion such as how long the participant has maintained
a ranking of poor, or other suitable criterion.
[0141] As used herein and in the claims, a "participant wheel" is
an ordered sequence used in assigning specific events to market
participants as a function of their behavior. In this embodiment,
there are four wheels: wheel LA is used to assign new stock loans
to lenders in a loan auction, wheel BA is used to assign new stock
loans to borrowers in a loan auction, wheel LR is used to assign
stock returns to lenders, and wheel BR is used to assign stock
recalls to borrowers. In other embodiments, other wheels may be
used.
[0142] Generally, a wheel is associated with a wheel formula,
specifying the number of appearances in the ordered sequence that a
participant earns in accordance with the participant's rank. For
example, a rank of "good" earns three appearances, a rank of
"normal" earns two appearances, and a rank of "poor" earns one
appearance.
[0143] At step 356, ELMS 200 determines the wheel appearances based
on the participant rankings
[0144] At step 358, ELMS 200 places the appearances into an ordered
sequence. In this embodiment, a pseudo-random sequence of numbers
corresponding to the number of appearances is generated, and then
the appearances are ordered according to the pseudo-random
sequence.
[0145] An example of wheel construction will now be discussed.
[0146] Assume that the participants for ELMS 200 are lenders L1,
L2, L3 and borrowers B1, B2, B3, B4 having ranks, determined as
above, shown in Table 3.
TABLE-US-00003 TABLE 3 lender participant rank borrower participant
rank L1 good B1 good L2 normal B2 normal L3 normal B3 normal B4
poor
Further assume that the formulas for the auction wheels and recall
and return wheels are as shown in Table 4. Note that for wheel BR,
a rank of "good" corresponds to zero appearances, that is, a
borrower participant with a rank of good will never experience a
stock loan recall. In other embodiments, other formulas are
used.
TABLE-US-00004 TABLE 4 wheel LA wheel LR wheel BA wheel BR rank
appears rank appears rank appears rank appears good 3 good 1 good 3
good 0 normal 2 normal 3 normal 2 normal 1 bad 1 bad 5 bad 1 bad
3
[0147] For wheel LA, L1 has a rank of good and thus three
appearances, denoted as L1a, L1b and L1c. Table 5 shows the
appearances in each wheel.
TABLE-US-00005 TABLE 5 no. appears wheel LA wheel LR wheel BA wheel
BR L1a, L1b, L1c L1a B1a, B1b, B1c L2a, L2b L2a, L2b, L2c B2a, B2b
B2a L3a, L3b L3a, L3b, L3c B3a, B3b B3a B4a B4a, B4b, B4c total 7 7
8 5
[0148] Wheel LA has seven appearances. ELMS 200 places the numbers
one through six in pseudo-random order, for example: 3462715.
Similarly, wheel LR has seven appearances, and ELMS 200 generates
the following pseudo-random sequence: 7143256; for wheel BA, the
sequence is 27361458, and for wheel BR, the sequence is: 41352.
[0149] Finally, ELMS 200 orders the appearances in accordance with
the pseudo-random sequence, as shown in Table 6. For example, wheel
BR is initially populated with five entries (B2a B3a B4a B4b B4c),
corresponding to a 1/5 chance of being chosen for each of
participants B2 and B3, and a 3/5 chance of being chosen for
participant B4. Consider entries (B2a B3a B4a B4b B4c) as having
the sequence (1 2 3 4 5). Now, the sequence digits are
pseudo-randomly arranged into the order 41352, corresponding to the
sequence B4b B2a B4a B4c B3a.
TABLE-US-00006 TABLE 6 (original sequence that populates the wheel)
wheel pseudo-random no. ordered sequence of participant appearances
LA 3462715 (L1a L1b L1c L2a L2b L3a L3b) L1c L2a L3a L1b L3b L1a
L2b LR 7143256 (L1a L2a L2b L2c L3a L3b L3c) L3c L1a L2c L2b L2a
L3a L3b BA 27361458 (B1a B1b B1c B2a B2b B3a B3b B4a) B1b B3b B1c
B3a B1a B2a B2b B4a BR 41352 (B2a B3a B4a B4b B4c) B4b B2a B4a B4c
B3a
[0150] Examples using the wheels are provided below.
[0151] At step 372, ELMS 200 sorts the stock loan inventory offers
and stock loan requests by security and price, grouping together
all offers and requests for a particular stock at a particular
price.
[0152] At step 373, ELMS 200 determines how much to match. In the
present embodiment, the full amount of each borrower's request can
be matched to one lender's inventory offer. In other embodiments,
constraints are applied, such as (i) a maximum of 50% of a lender's
inventory offer can be matched to one borrower, (ii) the portion of
a lender's inventory that can be matched to one borrower is limited
to three times the lender's inventory offer divided by the total
inventory being offered by all lenders, (iii) borrower requests are
divided into sub-requests each having a maximum of 20,000 shares,
and so on. The constraints exist to protect the exposure of
borrowers and lenders, and to ensure many transactions so that the
probability of loan matches more closely tracks what is expected
from the wheel.
[0153] At step 374, for each stock and price grouping, ELMS 200
takes the top borrower from wheel BA and the top lender from wheel
LA, and attempts to match the borrower's request with the lender's
inventory. If the lender's inventory amount is greater than or
equal to the borrower's request, then there is a full match. If the
inventory is less than the request, then there is a partial
match.
[0154] At step 376, ELMS 200 determines whether the borrower's
request is fully matched. If so, processing continues at step 380.
If not, at step 378, ELMS 200 takes the next lender from wheel LA,
and attempts to match the borrower's request with the lender's
inventory. If the lender's inventory amount is greater than or
equal to the borrower's request, then there is a full match. If the
inventory is less than the request, then there is a partial match.
Processing returns to step 376.
[0155] At step 380, the borrower's stock loan request has been
fully filled by matching to lender's inventory. The appearances,
taken from the wheels, used in the match are moved to the bottom of
the respective wheels. Eventually, as more auctions occur, these
appearances will rise to the top and again experience matches.
[0156] At step 382, ELMS 200 determines whether there is another
borrower in the auction by taking the next top ranked borrower on
wheel BA. If not , processing is complete. If so, at step 384, ELMS
200 determines whether there are more lenders with suitable
inventory; if so, processing returns to step 374. If not,
processing is complete.
[0157] An example of an auction match using the wheels LA and BA is
now discussed.
[0158] Assume that after step 372, ELMS 200 has created a grouping
for stock XYZ at loan price 2% as follows: Lender L1--100,000
shares, Lender L2--50,000 shares, Borrower B2--10,000 shares,
Borrower B3--20,000 shares. That is, Lender L1 is offering to lend
100,000 shares of XYZ at a loan price of FF-2%, Borrower B2 is
requesting a stock loan of 10,000 shares of XYZ at a loan price of
FF-2%, and so on.
[0159] At step 374, ELMS 200 takes the top borrower from wheel BA
and the top lender from wheel LA, and attempts to match the
borrower's request with the lender's inventory. The top borrower on
wheel BA corresponds to the first appearance on wheel BA, namely,
"B1b" as shown in Table 6. This is an appearance for borrower
B1.
[0160] However, borrower B1 is not part of the grouping for this
auction, so ELMS 200 continues to the next appearance on wheel BA,
namely, "B3b" as shown in Table 6. This is an appearance for
borrower B3 who is part of the grouping for this auction. So, ELMS
200 will now determine who provides the inventory for borrower B3's
loan request of 20,000 shares. ELMS 200 reads wheel LA and obtains
"L1c" as shown in Table 6. This is an appearance for lender L1, who
is part of the grouping for this auction, and is offering 100,000
shares for loan. ELMS 200 matches B3's request for 20,000 shares
against L1's inventory to create a stock loan of 20,000 shares of
XYZ at a rate of FF-2%.
[0161] At step 376, ELMS 200 determines that B3's request is fully
matched.
[0162] At step 380, ELMS 200 moves the appearances that
participated in the match to the bottom of the wheel, so that the
new sequence is as shown in Table 7.
TABLE-US-00007 TABLE 7 wheel pseudo-random no. ordered sequence of
participant appearances LA 3462715 (L1c L2a L3a L1b L3b L1a L2b)
L2a L3a L1b L3b L1a L2b L1c BA 27361458 (B1b B3b B1c B3a B1a B2a
B2b B4a) B1b B1c B3a B1a B2a B2b B4a B3b
[0163] At step 382, ELMS 200 determines that there is another
borrower in this auction, namely, borrower B2 wanting a stock loan
of 10,000 shares of XYZ. There are no other remaining borrowers, so
B2 is selected by obtaining its topmost appearance on wheel BA,
namely "B2a" from Table 7.
[0164] At step 384, ELMS 200 determines that there are more lenders
with suitable inventory, namely, L1 with 100,000-20,000=80,000
shares, and L2 with 50,000 shares, so processing returns to step
374.
[0165] At step 374, ELMS 200 reads wheel LA and obtains "L2a" as
shown in Table 7. This is an appearance for lender L2, who is part
of the grouping for this auction, and is offering 50,000 shares for
loan. ELMS 200 matches B2's request for 10,000 shares against L2's
inventory to create a stock loan of 10,000 shares of XYZ at a rate
of FF-2%.
[0166] At step 376, ELMS 200 determines that B2's request is fully
matched.
[0167] At step 380, ELMS 200 moves the appearances that
participated in the match to the bottom of the wheel, so that the
new sequence is as shown in Table 8.
TABLE-US-00008 TABLE 8 wheel pseudo-random no. ordered sequence of
participant appearances LA n/a (L2a L3a L1b L3b L1a L2b L1c) L3a
L1b L3b L1a L2b L1c L2a BA n/a (B1b B1c B3a B1a B2a B2b B4a B3b)
B1b B1c B3a B1a B2b B4a B3b B2a
[0168] In this example, all loan requests were fully matched.
However, in other examples, there might not be enough inventory to
satisfy all the loan requests, so borrowers with high ranks would
be more likely to get matches, whereas borrowers with low ranks
would be less likely to get matches.
[0169] As will be appreciated, lenders and borrowers with high
ranks have better chances of participating in auctions due to their
increased number of appearances on the wheels LA and BA.
[0170] In this embodiment, financial incentives are not part of an
auction match. However, in other embodiments, financial incentives
are provided. In one embodiment, hard to find securities can earn
incentives for lenders who provide them. In another embodiment,
borrowers who request loans greater than a threshold (number of
shares or value of loans) can earn incentives, generally
corresponding to the prior art practice of paying for order flow.
The incentives may be reductions in transactional usage fees for
ELMS 200 or other suitable type of incentive.
[0171] FIGS. 9A-9B are a flowchart showing automated incentives for
a borrower.
[0172] At step 400, EP 251 initiates a share return. At step 405,
ELMS 200 receives the share return.
[0173] At step 410, ELMS 200 allocates the share return to a
lender, discussed below with regard to FIG. 9B. At step 415, PLP
230 receives the share return allocation.
[0174] Since the share return is based on the ranking, this
demonstrates why lenders have an incentive to have a good
ranking.
[0175] At step 420, ELMS 200 determines whether EP 251 should be
financially incentivized for this return activity. In the case of a
return, in some embodiments, a borrower can earn positive financial
incentives for providing a too-early return of hard-to-borrow
stock, relative to the borrower's profile, or for providing a
return after holding a stock loan for an extremely long period. In
some embodiments, such as a marketplace for term loans, a borrower
can earn negative incentives for providing a too-early return of
stock.
[0176] The negative financial incentive is the difference between
the rates in primary market 210 and secondary market 220, the
difference being defined as "economic neutrality", plus a marginal
incentive, to encourage borrowers to choose the correct market when
the loan is initiated, or discourage borrowers from choosing the
wrong market.
Incentive=.SIGMA.LoanValue*(SecondaryRate+MargIncentiveRate-PrimaryRate)-
/365 days of Loan
Generally, the loan rates are based on the Federal Funds (FF) rate
modified by a particular number of basis points (BP) with one basis
point=0.01%. Assuming a current FF rate=2%, then, for example,
PrimaryRate=FF-100 BP=2%-100 BP=1%
SecondaryRate=FF-50 BP=2%-50 BP=1.5%
MargIncentiveRate=5 BP
Thus, the daily loan value is multiplied by (1.5%+5BP-1%)=0.55%
times the number of days outstanding divided by 365.
[0177] The incentive must be assessed separately for each loan
terminated because stability varies by security.
[0178] The incentive may be a different amount for each type of
activity. For example, stock returns that are after only a small
number of days may have a negative incentive of 5 BP, while stock
recalls may have a negative incentive of 10 BP. The incentive may
itself be a function of other parameters. For example, the
parameter may be the number of days less than a negative threshold
for a profile.
[0179] ELMS 200 assesses stability as a percentile position within
the loan distribution for similar securities, and compares the
assessed position to a predetermined threshold, as described
below.
[0180] The number of days is assessed relative to the typical
behavior for that market (primary or secondary) and that security.
For example, assume that for the security XYZ, and calculated over
all loan participants, for loan terminations initiated by the
borrower, the average loan length is 6 days (.mu.=6) and the
standard deviation is 1 day (.sigma.=1) (see FIG. 6 curve AA), and
a negative incentive is earned if the loan is terminated sooner
than two standard deviations from the average, i.e., the negative
incentive (left side of the curve in FIG. 6) is for loan
terminations of length .mu.-2.sigma.=6-2*1=4 days or shorter. A
positive incentive (right side of the curve in FIG. 6) is due if
the loan is terminated after two standard deviations from the
average, i.e., the reward is for loan terminations of length
.mu.+2.sigma.=6+2*1=8 days or longer. Loans terminations in the
range of (.mu.-2.sigma.) to (.mu.+2.sigma.) days, i.e., 4-8 days,
are normal and incur no incentive.
[0181] So, if a loan is terminated after 3 days, then for each of
the three days, the loan incurs a negative incentive of the loan
value times 0.55%/365 * number of days outstanding, in addition to
the economic neutrality amount (see formula above).
[0182] If, at step 420, it is determined that a financial incentive
is not warranted, then processing is complete.
[0183] If, at step 420, it is determined that a negative financial
incentive is warranted, then at step 425, the borrower is debited
by the disincentive amount. At step 430, the lender is credited by
the disincentive amount. In some embodiments, step 430 is omitted
and the funds are used according to a procedure beyond the scope of
this document.
[0184] If, at step 420, it is determined that a positive financial
incentive is warranted, then at step 435, the borrower is credited
by the incentive amount. At step 440, the lender is debited by the
incentive amount. In some embodiments, step 440 is omitted.
[0185] FIG. 9B provides detail for step 410 of FIG. 9A.
[0186] At step 450, ELMS 200 identifies lenders who are eligible to
accept a stock return, that is, lenders having outstanding loans of
the stock symbol being returned at the same rate as the stock being
returned.
[0187] At step 451, ELMS 200 determines the portion to assign, in a
manner similar to that discussed with respect to step 373 of FIG.
8C.
[0188] At step 452, ELMS 200 assigns the return to the top lender
appearance on wheel LR that is eligible.
[0189] At step 454, ELMS 200 determines whether the stock return
has been fully assigned. If not, then at step 456, ELMS 200 gets
the next eligible lender appearance from wheel LR and assigns the
return, and processing returns to step 454.
[0190] At step 458, ELMS 200 moves the just assigned lender
appearances to the bottom of wheel LR.
[0191] Partial returns are processed similarly.
[0192] An example of a stock return using the wheel LR is now
discussed.
[0193] Assume that borrower B4 returns 10,000 shares of XYZ.
[0194] At step 450, ELMS 200 identifies lenders L2 and L3 as
eligible to accept an XYZ stock return as each of them have loaned
100,000 shares of XYZ stock through ELMS 200. Here, lender L1 is
ineligible as it never offered shares of XYZ for a stock loan
through ELMS 200.
[0195] At step 452, ELMS 200 assigns the return to the top lender
appearance on wheel LR that is eligible. As seen in Table 6, wheel
LR has a sequence of: L3c L1a L2c L2b L2a L3a L3b. The first
appearance is L3c, corresponding to lender L3 who is eligible, so
the 10,000 share return from borrower B4 is assigned to lender
L3.
[0196] At step 454, ELMS 200 determines that the stock return has
been fully assigned.
[0197] At step 458, ELMS 200 moves the just assigned lender
appearance to the bottom of wheel LR. Table 9 shows the adjusted
wheel LR.
TABLE-US-00009 TABLE 9 wheel pseudo-random no. ordered sequence of
participant appearances LR 7143256 (L3c L1a L2c L2b L2a L3a L3b)
L1a L2c L2b L2a L3a L3b L3c
[0198] FIG. 9C provides detail for step 420 of FIG. 9A and step 520
of FIG. 10A.
[0199] At step 470, ELMS 200 gets the profile for this type of
market participant in this tier.
[0200] At step 472, ELMS 200 compares the action with the profile
to determine if the action is outside tier-normal behavior, that
is, below the negative threshold or above the positive threshold.
In some embodiments, negative incentives for early returns are not
implemented as a matter of marketplace policy.
[0201] At step 474, ELMS 200 determines whether to provide an
incentive. If the action is inside tier-normal behavior, then no
incentive is provided and processing is complete.
[0202] If the action is outside tier-normal behavior, then at step
476, ELMS 200 computes a financial incentive.
[0203] An example of allocating a financial incentive is now
discussed.
[0204] In this example, borrower CLP 241 returns 10,000 shares of
XYZ having a price of $6 per share after two days. Assume incentive
rates as set forth above in the discussion of FIG. 7.
[0205] At step 470, ELMS 200 gets profile BB of FIG. 6 as the
proper profile, with a negative threshold set at seven days.
[0206] At step 472, ELMS 200 compares the action, a share return
after two days, with the profile, having a negative threshold of
seven days, to determine that the action is outside tier-normal
behavior, that is, the share return was made abnormally soon after
the stock loan was established.
[0207] At step 474, ELMS 200 determines whether to provide an
incentive. Since the action is outside tier-normal behavior, the
determination is positive.
[0208] At step 476, ELMS 200 computes a financial incentive. In
this example,
Incentive=.SIGMA.LoanValue*(SecondaryRate+MargIncentiveRate-PrimaryRate)-
/365,
summed over the number of days that the loan is outstanding
Incentive=(2 days)*(10,000 shares)*($6 per share)*(0.55%)/365
Incentive=$183
[0209] At step 425 of FIG. 9A, the borrower is debited by $183.
[0210] FIGS. 10A-10B are a flowchart showing automated incentives
for a lender.
[0211] At step 500, PLP 230 initiates a share recall. At step 505,
ELMS 200 receives the share recall.
[0212] At step 510, ELMS 200 allocates the share recall to a
borrower, discussed below with regard to FIG. 10B. At step 515, EP
251 receives the share recall allocation. Since the share recall
allocation is based on the ranking, this demonstrates why borrowers
have an incentive to have a good ranking.
[0213] At step 520, ELMS 200 determines whether PLP 230 should be
financially incentivized for this recall activity. In the case of a
recall, a lender can earn only negative financial incentives for
providing a too-early recall.
[0214] If, at step 520, it is determined that a financial incentive
is not warranted, then processing is complete.
[0215] If, at step 520, it is determined that a negative financial
incentive is warranted, then at step 525, the lender is debited by
the disincentive amount, which may be calculated as a daily
interest rate times shares recalled for a number of days that would
make the recall be after a suitably long period, or the
disincentive may be calculated according to another suitable
procedure. At step 530, the borrower is credited by the
disincentive amount. In some embodiments, step 530 is omitted and
the funds are used according to a procedure beyond the scope of
this document.
[0216] FIG. 10B provides detail for step 510 of FIG. 10A.
[0217] At step 550, ELMS 200 identifies borrowers who are eligible
to accept a stock recall, that is, borrowers having loans of the
stock symbol being recalled.
[0218] At step 551, ELMS 200 determines the portion to assign, in a
manner similar to that discussed with respect to step 373 of FIG.
8C.
[0219] At step 552, ELMS 200 assigns the recall to the top borrower
appearance on wheel BR that is eligible.
[0220] At step 554, ELMS 200 determines whether the stock recall
has been fully assigned. If not, then at step 556, ELMS 200 gets
the next eligible borrower appearance from wheel BR and assigns the
recall, and processing returns to step 454.
[0221] At step 558, ELMS 200 moves the just assigned borrower
appearances to the bottom of wheel BR.
[0222] Partial recalls are processed similarly.
[0223] An example of a stock recall using the wheel BR is now
discussed.
[0224] Assume that lender L2 recalls 10,000 shares of XYZ.
[0225] At step 550, ELMS 200 identifies borrowers B1, B2, B4 as
eligible to accept an XYZ stock recall as each of them has a stock
loan of 5,000 shares of XYZ stock at the appropriate rate obtained
through ELMS 200. Here, borrower B3 is ineligible as it is not
borrowing shares of XYZ through ELMS 200.
[0226] At step 552, ELMS 200 assigns the return to the top borrower
appearance on wheel BR that is eligible. As seen in Table 6, wheel
BR has a sequence of: B4b B2a B4a B4c B3a. The first appearance is
B4b, corresponding to borrower B4 who is eligible, so the 10,000
share recall from lender L2 is assigned to borrower B4.
[0227] At step 554, ELMS 200 determines that the stock return has
not been fully assigned, since borrower B4 could accept a recall of
only 5,000 shares, the maximum it has borrowed. So, at step 556,
ELMS 200 gets the next appearance on wheel BR, namely, "B2a"
corresponding to borrower B2 who is eligible. ELMS 200 assigns the
remaining 10,000-5,000=5,000 shares to borrower B2.
[0228] At step 554, ELMS 200 determines that the stock return has
been fully assigned.
[0229] At step 558, ELMS 200 moves the just assigned borrower
appearances to the bottom of wheel BR. Table 10 shows the adjusted
wheel BR.
TABLE-US-00010 TABLE 10 wheel pseudo-random no. ordered sequence of
participant appearances BR 41352 (B4b B2a B4a B4c B3a) B4a B4c B3a
B4b B2a
[0230] Although an illustrative embodiment of the present
invention, and various modifications thereof, have been described
in detail herein with reference to the accompanying drawings, it is
to be understood that the invention is not limited to this precise
embodiment and the described modifications, and that various
changes and further modifications may be effected therein by one
skilled in the art without departing from the scope or spirit of
the invention as defined in the appended claims.
* * * * *
References