The Real Source of Trading Success
We know from research that persistent success at day trading does exist, and we also know that it is rare. In a sense, that should not surprise us: by definition, elite talent is rare. What makes trading difficult is that, in many fields, we can make a living from our work even if we are not operating at elite talent levels. In day trading, as well as other forms of trading, making a consistent living from trading *is* elite talent.
To illustrate some of trading's challenges, I ran a few numbers using SPY as my proxy trading instrument. From 2012 to the present, SPY has gained a little over 60 points. About half of those points were gained during overnight hours; half during U.S. day hours. The traditional day trader has limited himself/herself to about half the directional opportunity by trading only during U.S. hours.
Should traders seek greater directional opportunity by extending the time frame, they will find that the correlation between overnight changes and day session changes in the stock market since 2012 has been -.01. They might as well be totally different markets. If the trader extends to more of a swing time frame, the correlation between today's price change and tomorrow's since 2012 has been -.02. In other words, overall, what happens in the market during one short-term period offers no information about what will happen in the next period. We like to think directionally, in terms of trends, but--overall--what we can extrapolate from current markets to future ones is quite modest.
I have tested many patterns in markets and can attest that a correlation between market predictors and future price change that exceeds .20 is something quite special. Yet even that correlation implies that the predictors account for only 4% of future price movement. Our error variance is very high relative to what we can predict, even for statistically significant research.
That, of course, leads some to create models of sufficient complexity that they will promise far higher levels of predictive accuracy. As Derman notes, such complexity comes at the cost of fragility: modeling the financial world is fundamentally different from the modeling of the physical world. The laws of physics don't change readily. The behavior of market participants does. One of the better predictors of market bottoms in recent times was elevated volatility, particularly the "pure volatility" measure I have written about. During this most recent market decline, volatility blew out: what had been significant levels of volatility for calling market bottoms no longer applied to the new regime. We can calculate the odds of a given backtest being overfit, and it doesn't take much in the way of complexity to get to that point.
So where does that leave us? Simple patterns do not provide reliable profitability, breathless claims such as the above graphic notwithstanding. Complex patterns are all too likely to be overfit, producing great backtests but failing in real time performance. No, the answer is not to be simple or complex; the answer is to be different. Of the traders I worked with a decade ago, fewer than 5% are currently trading and experiencing success. In each case, they are doing something very different from what the standard trading books describe. They have found sources of "edge" in markets that they have made their own, and they have been consistent in exploiting those edges.
One trader, for example, came up with an ingenious method for identifying when trading in a particular asset was becoming highly crowded. He then looked for indications of loss of price momentum and took the other side of the crowded trade, benefiting from the herd running for the doors. A big part of his edge was that, if he didn't find the right patterns of crowding, he did not trade. He only played the game when the odds were on his side.
A very different trader at a financial institution obtained information from satellites that provided information about weather and crop planting patterns and used those data to predict yields for agricultural commodities. When markets were mispriced relative to the new data, trades with an edge could be placed.
Still another trader found that market moves at certain times of day had more likelihood of reversing than at other times of day, as different participants impacted the market throughout the day. By tracking the level of participation and movement and segmenting time differently from other traders, he was able to identify profitable trading patterns.
In the trading world as in the business world, "me too" is not a formula for success. The successful entrepreneur is the one who operates with a vision, doing something differently--and better--than rivals. It's not enough to plan your trade and trade your plan. Those plans have to be grounded in insight and unique information if they are to lead to ongoing success.
Further Reading: Keys to Day Trading Success
.
Comments
Post a Comment