Hedge Fund Market Wizard

Edward Thorp: The Innovator

For Edward Thorp, both gambling and investing are quite similar because both are based on probability and finding the edge can be done analytically. 

 

Thorp flouted the basic prejudices in both investing and gambling. He applied similar kinds of scientific and mathematical reasoning to each venture. The same principles and risk management is relevant to the trading lessons that he used in the markets. 

 

Winning against a roulette was earlier considered impossible mathematically. Thorp focussed on predicting the most probable ending zone for the ball. This different approach helped him create an edge of 44%. The comprehensive lesson is that what seems impossible sometimes is entirely possible if addressed with a different approach. This applies to trading as well. If the market cannot be beaten with conventional strategies, it is better to try unconventional ones.

 

Another important method apart from entry is altering the position size. The sizing should be smaller for lower probability trades and larger for higher probability ones. This can change a losing strategy into a winning one. 

 

Confidence is applicable not only in defining the trade size but also on reasonable risk management. It is sound advice to bet within your comfort zone. Emotions influence trading decisions and prove to be deadly. 

 

Many traders believe that there is a single remedy to define market behaviour and continuously try to find that Holy Grail of trading strategies. If it was true, then trading would become as simple as operating a money machine. However, there is no such single solution. For being successful, traders have to adapt themselves to changing market situations. It is not a fixed process but a dynamic one. 

 

The initial concept of Thorp’s approach was to balance long stock positions that had declined the most with short stock positions that had advanced the most to ensure that the portfolio’s net exposure was close to zero. However, when the return/risk started to erode, he shifted to a variant strategy that incorporated sector neutrality to market neutrality. When sector-neutral arbitrage also started losing its edge, he switched to another variant that neutralized the portfolio to the various factors. By the time the third repetition was altered, the original version had degraded considerably. This continuous adaptation and alteration helped Thorp to maintain a strong and consistent return/risk.

 

The Kelly criterion by John Kelly (a mathematical formula that helps investors and gamblers calculate what percentage of their money they should allocate to each investment or bet) can be useful to determine the optimal size of a trade. It can be mathematically illustrated and will generate a higher return in the long run than any other strategy. However, this model assumes that the winning probability and the ratio of profit/loss per bet are literally known. This assumption holds true for games but in trading, winning probability is not known and can be guessed with a broad degree of error. 

 

There is an upright penalty for overestimating the probability and ratio, when the Kelly criterion is used. The negative impact of overvaluing the size is twice as large as the negative impact of underrating it. Hence, if the accurate probabilities of winning are not known then the bet size should be significantly smaller than the full Kelly amount.

 

Moreover, even if the assumption is correct, the resulting equity surge would be volatile beyond people’s comfort level. The high volatility has important practical indications as well. Higher the volatility of the equity, the greater the chance of renouncing the approach during the decline.

 

Thorp suggests that even if the probability of winning can be estimated, using half Kelly is a better option. When the estimate is correct, even half the size will be a better option than full. However, if the estimation is subject to wide error, then even half size would be too large, and the Kelly criterion would be of limited use.

Did you like this unit?

1 0

Units 7/17