Remember me

Register  |   Lost password?

The Trading Mesh

Algorithmic Traders: Worthless Parasites or Riskless Arbitrageurs?

Fri, 08 Mar 2013 09:23:57 GMT           

By Bill Ulivieri
Cenacle Capital Management, LLC
www.cenaclecapital.com

 

Back in August 2012, Wired magazine ran a great article about High Frequency traders. (http://www.wired.com/business/2012/08/ff_wallstreet_trading/all/). Author Jerry Adler did an admirable job of describing the strategies and technologies used in high-speed algorithmic, computerized trading. He also made a good faith attempt of condensing the essential trading strategies of algorithmic traders, focusing on the technology side of the equation. In this article, Adler writes:  

 

“..under the “maker-taker” model, some exchanges offer tiny incentive payments, or rebates, for posting a quote (to buy or sell a stock) that results in a trade. The exchange charges the other side in the trade, the taker, a slightly higher fee and collects the difference. So an algo can buy a stock, earn a rebate, then sell the stock and earn a rebate for that too.”

 

A few months later, however, I was disheartened to read a blogger’s response to the article in the November 2012 issue of Wired magazine.  (Wired, November 2012; “Fast Cash”) 

 

The blogger opined that algorithmic traders are “worthless parasites.” Call me old fashioned, but in my opinion, algorithmic traders are more like riskless arbitrageurs than worthless parasites.

 

InvestorWord.com definition the term “riskless arbitrage” as:

 

“A risk-free transaction consisting of purchasing an asset at one price and simultaneously selling that same asset at a higher price, generating a profit on the difference.”

As an option “pit trader” at the Chicago Board Options Exchange for almost 30 years, I focused on trading strategies defined as “risk free.” One option strategy is called a box. A box, at its essence, is an interest rate play. A box is created by combining a bullish call option spread with a bearish put spread. The combined call spread and put spread net each other out in terms of risk. The notional value of a box is the difference between the option strike prices.

 

In 1987, if you bought a $500.00 box for $487.50 with 15 days left to expiration, you made $125 on a $500 investment over 15 business days. This equates to a 2.5% rate of return over 15 days, or roughly a 41% rate of return annualized. To keep the rate of return in perspective, Fed Fund interest rates were at 6.75%. This rate of return, of course, did not include trading commissions, exchange fees, exercise and assignment fees, monthly seat lease expenses, etc., etc.  A trader’s targeted annual gross rate of return was roughly six times the Fed Funds rate.

 

The same is true for other risk-free arbitrage strategies known as long conversion and short reversal. Pick a strike price of a pair of options, and take the bid and offer of the call and put markets along with the last price of the stock. Generally, if you traded one of the options on your market, the other two sides of the arbitrage were generally priced and quoted to lock in an attractive, risk-free rate of return. In a conversion strategy, you would sell call, buy put, and buy stock to enter into the trade. Generally, when a broker came into the pit with an order, it was which side of the market he “hit you with” that determined whether you were going to “leg” into the conversion or reversal strategy.

 

When pit traders execute a trading strategy, we always have to take a tiny “leg.” Taking a leg meant you assumed a little intraday risk until you could complete the other portions of the strategy to make it risk free. A leg could last seconds, minutes or hours, but almost always would be completed by the time the closing bell rang.

 

With Algorithmic Trading labeled as the “worthless parasite” of the trading industry, how fundamentally different is it than legging into a risk-free arbitrage strategy as pit trader?

 

In other words, traders would rarely buy or sell an option if we didn’t firmly believe the other sides needed to complete the leg were available. Boy, did we hit the other sides quickly! We were all in a high-speed race with other competing pit traders trying to lock in our risk-free hedge before it disappeared. The more participants on a trade, the less likely you could get the hedge off, which meant the less attractive the value of the trade was.

 

Algo traders buy stock at $7.01, sell it at 7.01 and collect the Exchange “make-take” rebate as an incentive for providing liquidity. Traders assumed risk, hedged the trade, and have no overnight position. Before expenses, they made $1.00 on a $700 investment in one trading day. Low priced, high volume stocks are popular feeding grounds for small algorithmic traders, trading 100,000 shares per day and will gross $800 to $1100 per day, not including execution costs, exchange connectivity, internet T1 lines, etc. 

 

Delta trading options and futures became more popular in the mid- to late-1980s and 1990s, because it allowed the trader the opportunity to trade hundreds of strike prices over an entire range of listed options during the day, and then cover any potential overnight gap risk by the end of the trading day.

 

Algo traders incarnate the Darwinian evolution of trading. Why are they considered a parasites, when back in the 1980’s, traders used Dell 386s to crank out Black Sholes option valuations for floor traders? Why is it different now?  A narrow bid/ask spread width is considered a financial benefit to institutional and retail investor alike. Adler partially agrees by stating that:

 

“…Conventional economics views all this as an unalloyed good: It is axiomatic that all trades are a net benefit to the economy because they enhance “liquidity,” the ability of investors to buy or sell assets at the best price. Indeed, in 2007 the SEC instituted an ambitious new rule, the national market system that opened the door to dozens of new venues for stock trading, but now that transaction times are measured in micro­seconds and prices are carried out to six decimal places, those opportunities have arguably gone past a point of diminishing returns.”

 

While today’s HF traders are buying and selling assets in a sub-10 millisecond world of exchange data, aren’t they really acting as arbitrageur’s by purchasing an asset at one price and simultaneously selling that same asset generating a profit on the difference?

 

Some believe it’s the Federal Reserve’s influence on the interest rate yield curve; mandated decimalization, and the law of unintended consequences which has nurtured the trading community to embrace high speed algorithmic trading. More likely, it’s just a manifestation of the same generation which brings us extreme sports, instant video streaming and an obsessive expression of individualism where a demutualized Exchange is considered a place to Occupy.

 

The opinions and writing contained in this article are of the author alone and do not necessarily represent those of HFTReview.com.

, , , , , , , , , , , , , , , , , , , , , , , , , , , , ,