Even a successful trader should hope for some good luck
Head of Equity Strategy
Summary: A strategy with a high 53% win ratio on daily trading commodity futures will on average realise strong annualised returns and world-class sharpe ratio. But the impact from randomness will create potential paths that will mask these facts and cause the trader to doubt the strategy and maybe give up or drift into other strategies.
On quiet Friday it is always worth reflecting a bit on the bigger issues in trading and investing. Today we are going to demonstrate why luck, or randomness if you will, is so important for even successful traders with a statistically significant edge. Our experiment will show why the market cleverly roots out the weak and ensures few will survive on a systematic approach.
What type of edge is required to be successful?
It all starts with a strategy, whether it is systematic or discretionary, which requires a positive edge. This is the fundamental foundation of speculation. No positive edge leads to loss of capital over time. But besides an edge, traders need good luck, but we will come back to that shortly. Let us start an experiment to build our chain of logic.
Let us assume that we have found a systematic approach (it could be some combination of technical indicators) that generates an expected win ratio of 53% by daily trading the 10 largest commodities futures. This is a pretty good edge on daily trading strategy in for the most part non-trending instruments. We sample with replacement 1,260 return observations (four years of trading) for each commodity based on its price history. We do this 1,000 times it generates many hypothetical paths our strategy. Based on these synthetic trading decisions we can create an equal-weight portfolio with total returns adjusted for roll-yield and trading costs (only commission and bid-ask spread, as we assume no price impact on trades). The plot below shows the distribution of annualized returns for this strategy over its 1,000 paths. The average annualized return is 19.7% which would take this trader into the premier league of trading.
Why luck is so important
The statistics on this strategy are so good with an average annualized sharpe ratio of 1.9 which is equivalent to this strategy marketing itself to prospective clients. Most would think that this is a clean race into the sunset with bank account ready to be filled. Unfortunately, randomness is the dominate factor in our lives and this goes for trading as well.
If we plot the wealth evolution of each path, we get the following plot. The black lines are all our 1,000 paths and the light blue line is the path for the single path at the median of all terminal wealth values. But the red circle is even more important. It shows that despite a strategy with a significant edge there are still 5% of all paths after three years that would break this trader. Maybe the threshold is even higher. There are indeed a bunch of paths that deliver less than 5% annualized and this trader would feel after four years that it is not worth all the stress and hard work. Why not just buy high-yield corporate bonds?
This shows why everyone needs some good luck and ideally a lot of it despite a world-class strategy. The market will shake so hard and pull traders through so many storms that even the best would have many periods of doubt. The three ingredients for trading are 1) find an edge, 2) hope for some good luck, and 3) persistence. The force of randomness is what gets traders to likely style-drift and thus never staying long enough with the strategy to see the edge crystalizing itself into attractive returns.
Because the sizeable impact from randomness we cannot truly know if the minor good and the really good are so because of skill or luck played a large role. Only traders and investors that have managed to stay in the arena for decades can truly be judged on their performance.
Please read our disclaimers:
- Notification on Non-Independent Investment Research (https://www.home.saxo/legal/niird/notification)
- Full disclaimer (https://www.home.saxo/en-gb/legal/disclaimer/saxo-disclaimer)