There are constant references to 2% position risk rule in social media and even in some websites and books. Most of the references are wrong and out of context.
The 2% rule was mentioned in a popular book many years ago as the risk management scheme used by a successful pit trader. I cannot know whether that was the first reference made to that rule so I will skip the name of the book here because anyway we care more about the essence and not about these details. The important fact is that this rule was subsequently generalized by various authors, traders and bloggers as many other rules that were taken out of context.
The rule is presented as follows:
“Never risk more than 2% of available capital on a single trade.”
This was probably a good rule for a certain style of trading in the past but as it happens usually, it was taken out of context and generalized by those who rush to turn everything that any successful trader says to a rule. In trading, fortunately or unfortunately, there are no general rules about anything.
Any rule of that sort, 2%, 1% even 0.5%, was maybe useful to pit traders, or now that pits are mostly history, to professional traders making many trades daily. When I traded large cap stocks for a hedge fund I used the 0.5% rule variant and I went as low as 0.2% during the 2008 bear market depending on number of open positions to reduce “portfolio heat.” But not every trader can use a position size rule of this sort since it implies tight stops for full utilization of trading capital. Let us see why below through examples. We will use the 2% rule without loss of generality.
Position size determination
A given level of risk per trade rule is useful when the position size can be calculated. This requires knowledge of a stop-loss level for the trades. The formulas for that are given in my free book Profitability and Systematic Trading.
But what about if the stop-loss level is not known or one is not used at all? Not all trading strategies involve stop-loss although they may use a tail risk stop. For example, let us consider the popular RSI(2) mean-reversion trading strategy in SPY with entry when RSI(2) < 10 and exit when RSI(2) > 70. We assume all trades are entered at the next open and commission is $0.01 per share.
Case 1: Fully invested capital – no 2% risk rule – no stop-loss
This backtest shows decent results from SPY inception although since last year performance has deteriorated. This is mainly due to a rise in momentum at the expense of mean-reversion. Sharpe is 0.62 for the strategy versus 0.52 for buy and hold in the same period.
Case 2: Fully invested capital – 2% stop-loss
How to use the 2% rule with this strategy? Obviously, if you try a 2% stop-loss, that will destroy the profitability, as shown below.
You do not want to use a 2% stop-loss with this strategy. Sharpe drops to 0.45 from 0.62, maximum drawdown increases and annualized return decreases significantly.
Note that the stop-loss must be increased to 20% for no adverse effect on performance! A large stop of this sort can be used for tail event protection (e.g. flash crash but usefulness is questionable even then) but it is not practical for position size determination.
But how to abide by the 2% rule touted by all those “gurus” and “experts”?
Case 3: Position size based on 2% risk rule and 4% stop-loss per trade
One way is to use the 2% rule to limit the impact of large losing trades. This strategy has 347 trades since SPY inception with 14 trades having a loss larger than 4%. Therefore, we will size the trades so that the maximum loss of equity is 2% per trade while the maximum loss per trade based on invested equity is 4%.
As expected, performance deteriorated further with annualized return dropping to 2.26% while Sharpe fell to 0.32.
One way of matching the performance in Case 1, is to increase the risk percent to 4%, remove the stop-loss and still size the positions accordingly. But this is equivalent to fully invested equity as in case 1. For relevant formulas subscribers to our premium service may refer to this article.
Is there a way of applying the 2% rule in the case of RSI(2) strategy without jeopardizing performance? The answer is No. The rule is not general, it was never meant to be, it has specific use for specific type of strategies and styles of trading. Anyone who refers to the 2% as general rule is probably confused and in some cases has little or no trading experience but just repeats what heard elsewhere.
Another example where the use of risk percent position size rule does not work is trend-following strategies. Effects are similar in terms of impact as in the RSI(2) case.
So what to do in case the 2% rule or equivalent cannot be used?
I go into more detail about this problem in Chapter 7 of my book Fooled by Technical Analysis: The perils of charting, backtesting and data-mining.
In a nutshell, for strategies where the 2% rule does not apply and equity must be fully invested to maintain profitability, traders may control risk by controlling the allocation of capital to the strategy. In the case of riskier strategies such as the RSI(2) mentioned above that make no use of stop-loss, the allocation may be low, in the order of 5% to 10% of available trading capital. Stop-loss outside the historical levels of maximum intraday trade loss can be also used for protection from tail events. There are other possible schemes that include buying protection in the form of options but those cost and also affect performance.
There are many more self-proclaimed experts and fake gurus in the trading space than people with skin-in-the-game. The constant reference to 2% position size rule as general without any specific qualifications is a proof of this. The 2% rule, or any x% rule of this sort, applies to frequent trading at low cost where the effect of a reduction in expectation due to position sizing is counterbalanced by an increase in frequency of trades.
The 2% (or X%) rule can be used with
- Medium frequency trading (scalping, intraday, short-term)
- Strategies with well-defined stop-loss levels
- Multiple open positions in different markets (portfolio heat control)
The 2% (or X%) rule cannot be used with
- Most mean-reversion strategies
- Medium to longer-term trend-following
- Strategies that do not use stop-loss
Alternative schemes for controlling risk when the rule does not apply must be used as mentioned above including but not limited to varying the strategy allocation accordingly.
If you found this article interesting, I invite you follow this blog via any of the methods below.
If you have any questions or comments, happy to connect on Twitter: @mikeharrisNY