The Kelly Criterion: The Intellectual Precursor to Optimal f
The Black Book of Day Trading Strategies
1,000 complete strategies · 31 chapters · Full trade plans
Introduction
To fully appreciate the significance of Ralph Vince's Optimal f, it is essential to understand its intellectual lineage, which can be traced directly to the groundbreaking work of John Kelly Jr. and the development of the Kelly Criterion. The Kelly Criterion, a formula born not in the halls of finance but in the research labs of Bell Telephone, laid the mathematical foundation for the modern theory of position sizing. This article explores the history, mathematics, and application of the Kelly Criterion, revealing its profound influence on the development of Optimal f and its enduring relevance in the world of professional trading and investment.
The Genesis of the Kelly Criterion
In 1956, John Kelly Jr., a researcher at Bell Labs, published a paper titled "A New Interpretation of Information Rate." The paper, which was intended to address a problem in long-distance telephone signal noise, introduced a formula for determining the optimal amount of capital to wager on a positive expectancy bet. This formula, which came to be known as the Kelly Criterion, was a revolutionary concept that had far-reaching implications beyond the field of information theory.
The Kelly Criterion is based on the principle of maximizing the geometric growth rate of a bankroll over the long run. For a simple binary outcome (win or lose), the formula is:
f* = (p * b - q) / b
f* = (p * b - q) / b
Where:
- f* is the optimal fraction of the bankroll to wager.
- p is the probability of winning.
- q is the probability of losing (1 - p).
- b is the payout on a winning bet.*
The Kelly Criterion in Action: A Gambling Example
To illustrate the power of the Kelly Criterion, consider a simple coin-tossing game where the coin is biased to land on heads 60% of the time. The payout for a correct bet on heads is 1:1. In this case, p = 0.6, q = 0.4, and b = 1. The Kelly Criterion would be:
f* = (0.6 * 1 - 0.4) / 1 = 0.2
f* = (0.6 * 1 - 0.4) / 1 = 0.2
This means that the optimal amount to wager on each toss is 20% of the bankroll. Any amount greater or less than 20% will result in a lower long-term geometric growth rate.
From Gambling to Investing: The Application of the Kelly Criterion to Financial Markets
The application of the Kelly Criterion to financial markets is not as straightforward as its application to a simple coin-tossing game. This is because financial markets are characterized by a continuous distribution of returns, where wins and losses are of varying magnitudes. However, the core principle of maximizing geometric growth remains the same.
To apply the Kelly Criterion to trading, we need to modify the formula to account for the variable nature of returns. This is where the concept of the "Kelly Ratio" comes into play. The Kelly Ratio is a measure of the risk-adjusted return of a trading system, and it is calculated as:
Kelly Ratio = (Win Rate * Average Win - Loss Rate * Average Loss) / (Average Win * Average Loss)
Kelly Ratio = (Win Rate * Average Win - Loss Rate * Average Loss) / (Average Win * Average Loss)
The Kelly Ratio can then be used to determine the optimal fraction of capital to risk on each trade.
The Limitations of the Kelly Criterion and the Rise of Optimal f
While the Kelly Criterion was a significant advancement in the theory of position sizing, it has several limitations that make its direct application to trading problematic. The most significant of these is its assumption of a known and stable distribution of returns. In reality, the distribution of returns in financial markets is constantly changing, which can make it difficult to accurately estimate the inputs for the Kelly formula.
This is where Ralph Vince's work on Optimal f introduced a revolutionary enhancement. Optimal f extends the core principle of geometric growth maximization to a framework that accommodates the variable and non-binary nature of trading returns. It does this by using a historical sequence of trades to find the fixed fraction of capital that would have maximized the geometric mean of returns over that period. This makes Optimal f a more robust and practical tool for position sizing in the real world of trading.
A Comparative Table
| Feature | Kelly Criterion | Optimal f |
|---|---|---|
| Core Principle | Maximizes geometric growth. | Maximizes geometric growth. |
| Input Data | Probabilities and payouts. | Historical trade data. |
| Assumptions | Known and stable distribution of returns. | Past performance is indicative of future performance. |
| Practicality | Difficult to apply directly to trading. | More practical for real-world trading. |
Conclusion
The Kelly Criterion was a revolutionary concept that laid the mathematical foundation for the modern theory of position sizing. While its direct application to trading is limited by its simplifying assumptions, its core principle of maximizing geometric growth has had a profound influence on the development of more sophisticated methodologies, such as Optimal f. By understanding the history, mathematics, and limitations of the Kelly Criterion, traders can gain a deeper appreciation for the power and elegance of Optimal f and its enduring relevance in the world of professional risk management.