FAQ: What do the risk statistics mean?

X
What do the risk statistics mean?
Click here to view the entire FAQ

The risk statistics are a broad indication of the risk and risk/reward profile of the trade results, rather than just their performance.

The risk/reward ratio is an expression of the system's performance in relation to its volatility (standard deviation). It prefers consistent results rather than wild swings. The figure is broadly analogous to an annualised Sharpe ratio, and figures in excess of +1 are notably good.

The risk of ruin is a Cox & Miller projection of the probability of a fall in the account balance based on the trading history, and particularly the standard deviation of results. The method estimates the likelihood of a fall in the balance at any time in the future. (N.B. Some investors question the applicability of Cox & Miller to trading, and prefer just to view the curve as an expression of volatility. It can yield very low estimates if volatility has historically been low.)

The spread of returns shows the number of daily/weekly % changes in the balance falling into each 1% band, e.g. the number of days on which the balance fell by 5%, or by more than 10%. Institutional investors will tend to prefer systems with a tightly clustered spread and few or no outliers. Systems with a wide spread will tend to have lower risk/reward ratios and a higher risk of ruin - because a small number of consecutive bad days/weeks would lead to major losses.

The deepest valley is the largest % fall in the balance from any peak to a subsequent trough. An investor who started trading at the top of the peak (rather than at the starting date of the results) would have seen this fall in their balance.

The loss from outset is the largest % fall which the system has seen relative to its initial deposit. It reflects the worst fall you would have seen relative to your original capital if you invested at the start of the system's results, but not necessarily if you started investing later (the "deepest valley").

 
Continue to site >