As the battle between federal regulators and Wall Street financiers rages on, a new player has entered the spotlight in the midst of the great debate over government’s involvement in the financial industry. High frequency traders, investors using sophisticated computer programs to rapidly buy and sell securities, have come into the crosshairs of federal authorities after several major incidents where malfunctions caused dramatic fluctuations in share prices on US and European indexes. Advocates for high frequency trading (HFT) argue that regulations impair the ability of financial institutions to produce profits and cut off an important source of liquidity for markets. European regulators have already proposed outright bans for HFT in their financial markets as America’s financial watchdogs look for answers to the question, “ban or regulate?” Either way, past glitches in trading software have exposed the volatility created by the excessive automation of HFT techniques undermining much needed investor confidence in US markets. Federal authorities must take immediate action to regulate HFT because of its susceptibility to errors in design and execution..
The Fall Of The Machines?
There are two main concerns regarding high frequency trading:
- Errors in the programming of high frequency trading systems have caused erratic behavior in the price movements of various asset classes in major stock indexes, increasing market volatility.
- Increased used of high speed trading methods has added false liquidity and the illusion of volume to securities that have a post-transaction net change near zero.
On the design error side, three incidents best demonstrate the risks of trading algorithms going haywire: Black Monday, the Flash Crash, and the rescue of Knight Capital Group.
The Monday That Lives In Infamy:
- One of the primary causes of the crash that began on October 19, 1987 was a trading method called “program trading”, Like HFT systems today, program trading platforms used computers to make investment decisions based on conditions given by the trader. It was common for institutional investors to place buy/sell orders for a stock or futures contract and then make the opposite call on the underlying stock or derivative. This was done to mitigate portfolio risk by setting up one position to offset potential losses from the other. The increased availability of computer technology made program trading incredibly popular around desks on Wall Street since traders could execute orders with greater accuracy and precision at desired price points.
- However, the advance in automation worked against them on that Monday when stock markets in Asia and Europe lost significant value prior to the opening of stock exchanges in the United States. Pre-market declines prompted a sell off on the Dow Jones Industrial Average which deepened once the pre-scheduled trades from program traders hit the market. The computers continued to sell into a depressed market, regardless of the amount of selling that was already happening, a cascading effect that resulted in a 508 point drop on the Dow.
Gone In 14 Seconds:
- Around 2:32 P.M. on Thursday May 6, 2010 high frequency trading computers at Waddell & Reed Financial processed a sale order for $4.1 billion worth of futures contracts, selling about 75,000 over 20 minutes of trading. Ordinarily, a sale of this magnitude is spread over five hours to maintain price stability as the value depreciates, but their algorithm was programmed to make the trades without respect to time or the sale price. Once the sale began, the trades were processed regardless of the severity of the declining asset values. High frequency trader in other sectors picked up on the selling activity and thanks to the sell off from Waddell, they were showing excess long positions so they started selling. The selling of outside investors prompted the trading computers at Waddell to sell even more aggressively; to the point where 200 contracts changed hands 27,000 times in just 14 seconds, about 1928 times a second.
- The situation worsened when the sell off in futures bled into the equities markets later in the trading day. Investors saw futures for cheap and bought into their sale, but turned and sold off cash shares on the stock exchanges (NYSE, NASDAQ etc.) When it was all said an done, the computer-based sell off in the futures exchanges caused extreme price movements in stocks as companies saw their share prices drop to pennies, or rise to nearly $100,000.00 a share. This volatility in prices continued until stabilizers in the futures exchanges halted trading for 5 seconds to reboot and return prices to appropriate levels.
If It Walks Like Bear and Talks Like Lehman:
- Knight Capital Group is a financial services firm that specialized in electronic market making, computerized trading systems, and electronic securities trading. During 30 minutes of the August 12, 2012 trading day, computers at Knight sold off all its investment positions from the previous day. Reports indicate that the trading errors weren’t stopped because of a lack of proper oversight capability on the part of Knight and the lack of a termination command for the rogue algorithm or other failsafe. By the time trading ended on Thursday, shares of Knight Capital lost 95% of their value (32% on Wednesday followed by another 63% that Thursday) settling to $2.58 a share. As a result of the trading error, Knight incurred a $440 million loss, far greater than its $289 million in revenue from the previous quarter, and the firm needed emergency capital to open for business the next day. A group of six financiers stepped in to provide Knight the resources it needed to stave off the complete collapse of the firm.
False Liquidity And The Semblance Of Volume
HFT gives securities ‘false liquidity’ in the sense that large numbers of trades, performed over a short period, would mimic an increase in market demand and achieve no significant increase in mobility. Remember from What Is Liquidity Risk, that liquidity describes how easily a security can be bought and then sold back on the market at a profit to the seller. High frequency trading uses computers to buy or sell large pools of securities in fractions of a second, which gives the traded asset the appearance of high liquidity. But this ‘liquidity’ isn’t real, because the net change in volume between the purchase and sale are too close to zero. It’s as if you were to go into the grocery store, buy 10 apples paying no tax, then return them and be compensated the tax by accident. In the end, you returned the exact number of apples you originally bought bearing no weight on the overall marketability of the apples.
It’s this parity that makes share volume a less useful metric. Looking back at What Is High Frequency Trading, these traders end their day with no net investment positions, meaning that they put back what they take out at the end of every trading day. Share volume is useful to see how many times a particular security is bought and sold, ideally to find a variance between the number of buyers/sellers and forecast where that security is headed. HFT heavy markets are full of transactions that have a 1:1 ratio, resulting in much less variance, because the number of sales by each investor would counterbalance their purchases.
How Could We Regulate HFT?
Based on past issues with computerized trading, any regulations created for this system would need the following elements:
- Regulatory agencies would need to establish uniform standards, to be used between each body, for what types of trading activities qualify as “high-speed” or “high frequency”. There also would need to be definitions qualifying traders and investors as “high frequency traders” and for types of computer software that perform these trades
- A comprehensive list of algorithms and software used for “high frequency trading”, which would update as the market for it continues to evolve, would be needed
- Regulators would be required to have an exhaustive list of firms that self-identify as “high frequency” or “high speed” institutions, and to use the above standards to qualify other financial institutions who participate in this type of trading.
- There needs to be a way for authorities to verify that firms using HFT systems have a fast, effective way of shutting down, or disabling their software in the event that the programs act unexpectedly or out of sync with the firms trading agenda that day.
- Also needed is a verification method to ensure that one firm’s computer can recognize erratic market behavior and stop its own trading to offset the “hot potato” effect seen in the Flash Crash.
Having uniform defitions for HFT would allow the various regulatory agencies involved to approach the issue from the same direction, moving in cohesion and creating rules that make sense across exchanges. Collecting data on the firms and the software they use makes it easier to regulate having a smaller footprint on the markets as whole. Regulators need to make rules, but the rules also need to be specifically tailored to the issue and have as little spill over as possible.
A Last Word
In light of events like the Flash Crash and Knight Capital, it’s more than clear that some combination of federal oversight and corporate responsibility is needed to mitigate the risks of HFT. Market confidence is a fragile thing, especially in the current climate, where concerns over Europe, tax policy, and the ‘fiscal cliff’ drive investors more than company fundamentals do. Errors from high frequency trading can exacerbate sell offs, drive markets into a spiral and take firms into insolvency in the blink of an eye. We must do more to support markets against these concerns and keep America’s exchanges a safe place for all comers to invest.