Capital Markets Being Undermined by New Technology
Examination of Current Trading Environment:
Recent techniques and tools using new technology are being used to gain unfair financial advantages in the trading environments of the world’s trading exchanges. Since it is the recognized goal of Exchanges, Governments and other security regulatory group to ensure fairness and promote investor confidence, a constant vigilance must be maintained to identify where changes can negative impact these environments and thus the confidence in which they are held.1 2
It has long been known that the speed of receipt, analysis, and action on information can be of significant help in formulating and executing financial transactions and strategies. The invention of the telegraph, semaphore, and even smoke and fire signals have been used throughout history to get market data as soon as possible so as to give a financial advantage to those that received such knowledge before other investors or property holders. In modern times, governments have sought to inject a degree of fairness into this process by proscribing “insider trading”, market manipulation, and by introducing various other rules and laws to level the “playing field” in information gathering, dissemination, and reliability. This results in greater confidence and participation of the public and therefore a larger market to provide capital for the needs of society as well as a broader participation by society as a whole in the fruits of the economy.
In testimony before the Senate Banking Subcommittee on Securities, Insurance, and Investment Mr. James A. Brigagliano of the SEC stated, “As markets evolve, the Commission must continually seek to preserve the essential role of the public markets in promoting efficient price discovery, fair competition, and investor protection and confidence.”3 Further he stated,
1
Investopedia, Definition of ‘Exchange’ http://www.investopedia.com/terms/e/exchange.asp#axzz1kz5xrqvf (Jan 30, 2012)
2 Word Federation of Exhanges, Qualifying Characteristics for WFE Member Exchanges
http://www.worldexchanges.org/files/file/Qualifying%20Characteristics%20for%20WFE%20Member%20Exchanges%20 -%20October%202010(1).pdf (Jan 20, 2012)
3
James A. Brigagliano, Testimony Concerning Dark Pools, Flash Orders, High Frequency
Trading, and Other Market Structure Issues , http://www.sec.gov/news/testimony/2009/ts102809jab.htm (Jan30,2012)
“An exchange brings together the orders of multiple buyers and sellers and is required to provide the best bid and offer prices for each stock that it trades, as well as last-sale information for each trade that takes place on that exchange. This information is collected and made public through consolidated systems that are approved and overseen by the SEC. Any investor in the United States can see the best quotation and the last-sale price of any listed stock, in real time. This transparency is a key element of the national market system mandated by Congress. “ A major problem is that in the current state of technology nothing is done in “real time”
Recent advances in computer technology and electronic communications have resulted in a new paradigm for investing through the major exchanges. While until recently buying and selling (trades) were done manually by human traders that would make and accept offers to buy and sell, now the vast majority of trades are done through the matching of offers and acceptances by computers within the control of the organized Exchanges. The speed of execution of trades is almost beyond comprehension. Trades are now executed routinely in under 200 microseconds (a microsecond is one millionth of a second)4. A recent analysis of the trading volume on the London Stock Exchange showed that over one trillion trades were completed in 2011 on just this one exchange.
A companion to computerized trading by exchanges are high speed systems offered by exchanges, to deliver reports of trading activity.. Data is offered on a trade by trade basis as fast as trades are executed. As each trade is executed a report of the trade is distributed to subscribing members telling the essentials of the trade so that the price and quantity involved can be analyzed by computer algorithms to supposedly anticipate trends and opportunities.
It is obvious that this speed of investing decision making is beyond what would be possible for humans. Instead investing decisions are made by computers generating instructions to other computers to execute trades based on information supplied by still other computers.5
Traders seeking to gain a trading advantage over other traders are utilizing many schemes to place trades and get trading information before competing traders. This requires faster and faster communication techniques.
A key factor in maximizing the speed of obtaining information and getting trade requests to the computers of the exchange is the “latency” of the data transmission. The speed of transmission while extremely fast is finite and thus is not in “real time”. The speed of sending an electronic message from sender to receiver is theoretically limited by the speed of the movement of electricity which approximates the speed of light. Even though this is extremely fast it is not instantaneous. Thus the physical proximity of the source of the transmission to the receiver of the transmission will influence the interval of delay (“latency”) between sending and
4 London Stock Exchange, Press Releases 2011 70 / 2011 (Jan 30, 2011) http://www.londonstockexchange.com/aboutthe-exchange/media-relations/press-releases/2011/endofyear2011.htm
Wikipedia, Algorithmic trading . http://en.wikipedia.org/wiki/Algorithmic_trading (Jan 30, 2012)
receiving. A whole discipline has been developed to shorten this latency and use this to gain advantage in trading environments.6 A competitive advantage has been sought and is being used at great expense by some trading firms in setting up “Collocation” of facilities. This means locating the computers of the traders in the same physical locale as the Exchange, or at someplace near a site set by the Exchange as a ‘portal’ or doorway to the communication link to the Exchange’s computers.7
Effects of current system
The use of current trading systems has resulted in a few traders gaining advantage over the rest of the investing public, especially the few traders that have access to “Ultra Low Latency”, High Speed Algorithmic Trading, Collocated Facilities, and direct access to trading centers’ trading computers. This creates a market that is slanted to the benefit of the few powerful participants to the detriment of the vast majority of the trading public8.
These few, exploit temporary deviations in prices shown on items offered on multiple markets. The advantages of such arbitrage trading facilitates gives unfair advantage to those using “ultra low level latency” connection strategy. Such techniques are actively used in all liquid securities, including equities, bonds, futures, foreign exchange, etc., which trade within multiple trading venues. Such strategies may also involve classical arbitrage strategies, such as covered interest rate parity in the foreign exchange market, which gives a relation between the prices of a domestic bond, a bond denominated in a foreign currency, the spot price of the currency, and the price of a forward contract on the currency. High-frequency trading allows similar arbitrages using models of greater complexity involving many more than four securities. The TABB Group estimates that annual aggregate profits of high-frequency arbitrage strategies currently exceed US$21 billion. This in effect is the garnering of profits from the rest of the investing community including pension funds and mutual funds.
Other questionable practices include “Flash Trading” where an order is generated without the intention of executing the order even if it has been accepted. In the “new” environment a computer generates an order then generates a cancel of that order within the time frame allowed for such cancellation. Note that if the “Flashed Order” was accepted this information can be seen by the person generating this order which was cancelled thus allowing the gaining of knowledge with out commitment 9
6 Ultra-Low Latency Direct Market Access (ULLDMA) (6,000+ Members) Group http://www.linkedin.com/
7
Ivy Schmerken in Wall Street & Exchanges 12/20/2011 issue, Exchanges Say Collocation Is A Regulated
Business http://www.wallstreetandtech.com/exchanges/232300847 (Jan 30, 2012)
8 Wikipedia, High Frequency Trading http://en.wikipedia.org/wiki/High-frequency_trading#Ticker_tape_trading (Jan 30, 2012)
9
TECHNICAL COMMITTEE OF THE INTERNATIONAL ORGANIZATION OF SECURITIES
COMMISSIONS, Regulatory Issues Raised by the Impact of Technological Changes on
page-3 Jan 30, 2012 jerry@fairtradingdevices.com (424)229-1623
It has been conjectured that “Flash Crashes” have been caused by such trading practices10.
Most Algorithmic traders claim trade secret protection in protecting their algorithms from disclosure to regulatory agencies as well as the exchanges themselves.
While it is doubtful whether, as a matter of public policy, this type of trading is healthy for our capital financial markets, even more disturbing are the possibilities for deliberately interfering with the flow of data to the advantage of a participant. The recognition of this possible weaknesses in such an environment can be even more alarming and possibly damaging to the market’s confidence. The topology and standard transmission techniques make it relatively simple for an unscrupulous participant, to interfere with traffic by injecting extra signals on the common communications path. The communication system used by the financial system is based on an Ethernet like model using CSMA/CD11. The information is transmitted on wire or optical fiber paths that are shared by many users. The communication protocol scheme depends on recognition of corrupted packets being caught, recognized by the receiver, reported to the sender and then retransmitted by the sender. If the packets are interfered with either deliberately or by just coincidence the latency of these trades are increased, because of the need to retransmit, thus benefiting the non corrupted trades.
If we examine the reality of the many collocated or almost collocated firms using techniques to “arbitrage” various scattered exchanges it is apparent that the availability of price data from the exchange located at the nearest point to the trader will be faster than the data from the distant exchange and thus to maximize the advantage of the opportunity one would have to in some way delay the information or trades to traders in closer proximity to the distant exchange. 12
Also disturbing are recent announcements concerning industrial espionage, hacking, malware attacks. The advances in high speed computing make it reasonable to assume that decryption of stock trades is possible so that “snooping” on trades within communication paths can be
Market Integrity and Efficiency CR02/11 July 2011
http://www.iosco.org/library/pubdocs/pdf/IOSCOPD354.pdf (Jan30, 2012)
10
Zachary A. Goldfarb, The Washington Post 10/1/2010, Report examines May’s ‘flash crash,’ expresses concern over high-speed trading http://www.washingtonpost.com/wpdyn/content/article/2010/10/01/AR2010100103969.html (Jan 30,2012)
11
Omnisecu.com, Comparison between TCP/IP and OSI http://www.omnisecu.com/tcpip/tcpip
model.htm (Jan 30,2012
12 Pavitra Kumar, Michael Goldstein, and Frank Graves ,The Brattle Group issue 02/2011, Trading at the speed of Light: The impact of High-Frequency Trading on Market Performance, Regulatory Oversight, and securities Litigation
http://www.futuresindustry.org/ptg/downloads/Upload932.pdf (Jan 30, 2012) (
successful. The enormous potential profits and low likelihood of detection make such activity probable.13
Is there a real problem?
Analysis of published data on TCP/IP connectivity and transmission confirms that the presently utilized infrastructure is vulnerable to snooping and interference which can lead to large profits for those taking advantage of the weaknesses. The presence of laws, regulations, and industry hardware standards would probably not significantly deter misconduct because of the huge profit potential, difficulty of proving the errant behaviour and less than certain penalties.
An appreciation of recent developments in co-location such as trading entities paying extremely high prices for close physical proximity to trading exchanges, the prevalence of extremely powerful, computer technology in these closely located trading centers all are clear indications of the probability that such behaviour is actually occurring.
Huge trading advantages could be gained by “snooping” on and/or delaying trading by other participants in the market. If, as is stated in publications, advantage can be gained by the reduction of latency of the trade transmissions, an even greater advantage can be obtained by the deliberate interference with trade transmissions thus even further magnifying the impact of trading transmission speed advantages gained by co-location. Under the normal Internet transmission to and from Exchanges some data collisions are expected. It is currently feasible to deliberately inject errors into packets within a node link of a network to force packet retransmission and thus in effect increase the latency delay of such packets sent by competitors. This activity can exist without it becoming obvious that such interference is occurring. Even if it occurs and suspected, it would be very difficult to prove the collisions were deliberately caused and to prove who is responsible. When a collision occurs whether from normal coincidence or deliberate injection, the corrupted packets must be retransmitted to complete the movement of a packet thus causing a delay in the reception of an actionable trade.
Further tremendous advantage could be gained if the persons or firms were able to actually read and make decisions on trades that were coming from others, and then delay these “cracked trades” by packet corruption. While it is assumed that current Internet activity is protected by encryption schemes, it is possible (even probable) that the encryption can be broken. Much activity in hacking and cracking encryption schemes is going on world wide. Because of the use of “phishing” schemes, “Man-inthe-Middle”, spoofing attacks, “Mal-ware” insertion, industrial espionage, and the
Home: Standards Initiatives: Public-Key Cryptography Standards (PKCS): PKCS #7: Cryptographic Message Syntax
Standard: 3.1 RSA, 3.1.3 What would it take to break the RSA cryptosystem?
http://www.rsa.com/rsalabs/node.asp?id=2216 Jan 30, 2011)
possibility of using super computers with new algorithms as may be invented, it would be naive to assume that encrypted transmissions are now, and will remain in the future immune from snooping. The U.S. government has for years proscribed the export of programs using encryption stronger than specified key lengths because it assumes the extremely fast computers needed to crack the commercially available key lengths are available. One must assume that governments have already developed cryptography cracking methodology. Recent press releases accusing China of breaching trade secrets of U.S.Firms attest to the seriousness and credibility of such activities.14
It is important to realize that almost all cryptography used today depends on the periodic exchange of “secrets”. In order to be useful to an eavesdropper the secrets do not have to be cracked in real-time. Usually the overall “asymmetric Master Secret” (identity certificate, Public/Private key pairs, etc.) will have an extended lifetime. Once that secret is known (or cracked) the communications between the individual ends of a packet exchange which will be passed using a symmetric secret (e.g. “session key”) that was communicated in an encoded form by using the “Master Secret”, are thus open to examination. If the time that it takes to “crack” the codes is less than the time the code is actually in use, The time interval (second, minutes, hours, or even days) that the cracked code is used to transmit the session key the communications would be exposed to snooping. The trades transmitted during this time would be open to the unauthorized listener. This can result in the unauthorized listener realizing large profits.
How do we fix the problem?
There needs to be fix to the problems outlined. The fix needs to be economical, easily implemented, interfere with trading as little as possible, be consistent with the stated objectives, be flexible to react to changing conditions and needs auditability. All this can be achieved by the introduction of an “Exchange Gateway Device” at each Exchange or trading center. An Exchange Gateway Device would screen all communications going to the Exchange’s trading computers. It would establish “fairness queues” which store and release trades and information at intervals, in blocks, and in sequences determined by the Exchange and/or oversight agencies to insure that an orderly and fair market is achieved. In other words this device can save all communications on a packet by packet basis. Examine each packet and determine if it was corrupted. It can store the packets in various queues. The contents of the queues would, periodically examined, arranged in a “fairness enhancing sequence” and released to the trading computers to be executed. The delay introduced need not be great (for instance it is possible that an ideal could be less than 1 second) and could be varied with different trading conditions,
Rob Iati in Advanced Trading July 10, 2009 issue, The Real Story of Trading Software Espionage
http://www.advancedtrading.com/algorithms/218401501 (Jan 30,2012)
but it would be long enough to ensure that latency and interference would not be a factor in profits realized. The queuing delays, releasing intervals, and “Fairness Sequencing Algorithms” would be open to public scrutiny and as empirical data is accumulated could be fine tuned to maximize benefit. The device could accumulate historical information for analysis to allow audit of the trades and frequencies.
Why an Exchange Gateway Device?
An Exchange Gateway Device, as I have conceived it, is offered as the solution to the outlined problems because it can be implemented quickly, at moderate cost, offer transparency, and does not require major changes to the communication or trading infrastructure or software. The parameters it uses to insure fairness and security can be changed in response to changing conditions and newly recognized needs. By offering a standardized means of monitoring and reporting it could allow oversight by regulators and industry groups which can reinforce the confidence and trust of the global investing community.
About the Author:
Jerome Simonoff is the President of CHAX, Inc. an independent software development company specializing in funds transfer software. He has been active in the information technology field as a consultant and entrepreneur for many years. He is the inventor of the underlying technology necessary to the Check 21 banking law enacted in 2003 (http://en.wikipedia.org/wiki/Check_21_Act ) – Patent US 7,945,084 B2. He sold this patent to Bank of America in 2008.
Leave your opinion!
You must be logged in to post a comment.