I don't really understand the question if I'm being honest. I assume that the distribution of forecasts is sufficiently nice and stationary such that using a fixed figure for mean absolute deviation is a good thing to do. This doesn't have to be Gaussian - any symmetric distribution that is stable is fine. I also assume that standard deviation and the correlation of the underlying instrument returns are appropriate and stable measures of risk. GAT
I'd certainly rather follow my own system than a random one on website! Having said that though, I could totally see myself give my cash over to another quant fund if I understood the strategy, as there's some hassle to running it yourself. I would never download a strategy from the internet and just run it blindly.
For any daily strategy, you need to backtest many years on many instruments to achieve statistical significance. It's too hard to tell from a demo account if the strategy is working.
Just to show that even the so called 'experts' make mistakes; I hadn't forseen that doing this would affect my margin usage. Over the last few days I've come up against margin limits on opening trades; something that would normally happen only once a year. Weighing up the options: add more cash to my account. I normally keep about 6 months of household expenditure in cash in a bank account. I could throw most of this cash at the margin problem, and then have a fraught existence where I tried to ensure that dividends from other investments and book royalties cover our bills every month. I'd rather not. I'm also opposed to the idea of adding cash to my investment account; even if the capital risk isn't theoretically increasing it feels like I'm putting more money at risk (an in practice I am exposing currently safe cash to the risk of a market shock or broker default). sell some stocks in my account to free up margin. This might result in a tax bill; I wouldn't know for sure until the end of the year as it will depend on whether my futures trading shows a taxable profit or not (which is why I don't normally sell stocks until the last few days of the year to optimise tax). Also selling stocks negates the point of taking the hedge off! reduce the gearing on my futures trading until I no longer have margin problems. This means permanently lowering my risk target. Implicit in this is that I've transferred risk from futures trading to long only equities (by removing my hedge). From a SR and diversification point of view that's probably bad. reinstate the hedge on my account Essentially by taking the hedge off I'd (a) increased the overall risk on my portfolio and (b) allocated that increase in risk entirely to equities. Whilst the latter might have made sense, the former did not. I've decided to put the hedge back on. For curiosity I'll calculate any gain or loss from the direct trading of the hedge; but in practice there will be other gains or losses from lost trading opportunity that I cannot be bothered to calculate. GAT
Closed at 3597, with E50 of slippage. So a profit, although as I said it's hard to know for sure. GAT
@globalarbtrader One more possibility is to analyse the average profit per instrument versus the amount of margin that instrument uses. Thus a sort of "bang for your buck" analysis. Although my account is a lot smaller than yours I found that it made sense to reduce the portfolio weight of certain instruments to avoid them using too much of the total available margin. It is a bit more subtle approach than just reducing the overall gearing, as you suggest in the third bullet.