One of the best aspects of futures is the high leverage which allows one to magnify their gains on even a small edge. But, that’s the ugly side too.
Scalping the ES with a 5 lot: you can be up $500 in a matter of minutes. Hit a few good trades and you can be up several thousands, and this is all possible on a tiny account.
The ugly is the high leverage makes it easy to fool oneself (and sometimes others) that you have an edge. Hit a few good trades and you could double a small account in a week.
One way to look at your performance then is in terms of multiples of per trade risk units, similar to R. The higher the multiple then the more confidence that one has an actual edge.
So, let’s say you risk $500 per trade on average and you’re up $5,000 then you would only be up 10 R Trade Units or 10x your average risk per trade.
The R factor in terms of daily loss units is also another useful way to look at return. So, let’s say in that case you risked $1,000 per day then you are up 5 R Daily Risk Units.
R is also useful for understanding what is possible to make in a day. Most scalps have a less then 1R return. A very solid scalp might return 2R. A good day swing trade may return 4R.
First, there are several specific factors that make translating a discretionary trading method into live trading more difficult or less difficult. Those factors are:
- Discretionary vs Systematic components. Methods that are more discretionary are more difficult to translate. The more things one can make less discretionary then the easier it will be to get a similar performance in live and sim.
- Frequency. Higher trade frequency makes small errors more likely to compound and introduces risk for discretionary trader because the higher frequency suggest that one must be operating in a systematic way. Discretionary methods that involve placing a lot of trades per day will be more difficult to translate.
- Leverage. Higher leverage can make mistakes way more expensive beyond what was anticipated.
- Risk. Obviously, if you’re risking less then it is easier to stay with a method when losing.
- Consistency. If a method isn’t very consistent but relies on scoring infrequent large trades it will be harder to trade.
- Inverse Risk Reward. Most scalping strategies require or do best with “inverse risk/reward”, i.e. risking more then you make. Strategies like this are more difficult to translate to live because you know that you can lose more then you stand to make on any give trade and a few losses can ruin a good win ratio.
- Correlated Loss Probability. This is a big risk.
- Robustness & Confidence. Strategies that have shorter histories do not inspire confidence.
- Number of losses per day. Strategies that require taking many losses per day will be difficult to trade regardless of profitability.
What’s the expert solution for minimizing the ugly?
- For trades that are highly discretionary then decrease trade frequency and decrease sensitivity by using options and/or larger stops.
- Use programmed setups with edge so that you know exactly what you are trying to do.
- Trading in the live is too slow to get enough experience with discretionary methods that rely on frequent trading. If you are day trading then you need to market replay a ton. I am not talking about replaying a few days and calling it good. Would you trust a strategy with a few weeks data?
Again this is why you need programmed setups because market replaying without programmed setups cannot inspire the same confidence– however, it can be used to develop market cognition for new systems.
So what’s the number here? Hard to say but most discretionary traders make most of their money in the first hours of the market. If you play back the first 2 hours of the day at 4x speed then you can replay a day in 30 minutes. You could replay a month in 10 hours, 3 months in 30 hours, 6 months in 60 hours.
- Look at your return in multiplies of average per trade R and daily R. While I do not have a formula, I think around 20 daily R would start to give some assurance. In addition, I think the CALMAR is a good measure: return/max dd. A CALMAR of 4x to 5x or higher would be desirable.