How do you have the same capital for when you are buying just one of them or both of them at same time? I don't understand that. Are you 2x leveraged when you buy the 2nd one? What about numerator question? Also, does the standard deviation (denominator) make sense? We need more information.
Because this is already built in to the markets, the thing about backtesting is one small detail, I know some consultants who built an architecture that runs everything from hedge funds to crypto exchanges to wealth management, they identified eight (8) key market (areas) where most entities such as a hedge fund only focus on one area, and you would think only one area would be active at any time but it turns out they are all active, however some areas wait for larger movements to capture everything underneath it, they stay dormant until the goose is fattened. Today everything is active at once, but everyone here will only be able to plan and trade on one of the areas meaning effectively one in eight successful trading days, more or less if they had one trade per day, that's why institutions have different divisions to track the different areas, the markets have already built in offsetting trades and more often than not individual stocks move up slowly and individually but go down all at once, and in each stock and also on each timeframe and on each fundamental each stock can be at a different level which takes a larger move to revert to the mean. Then the quetion is can you find a what level each stock is at, kind of but the amount of data is so vast that tracking on more than one instrument becomes a problem, even with technology, the simple outcome is you track one stock well and sporadically or track many constantly with ever decreasing returns compared to tracking one, but tracking one stock if you miss the entry/exit it will destroy your returns whereas with many if you miss the entry/exit it will have less effect on a lower return profile. Risk/reward is not just for placing the trades, it's for the entire strategy, what do I do, I have access to the fintech so can track all eight levels on a few assets, let's say it makes a little bit of a difference, their Sharpe ratio comes out above 3%!
The strategy was tested for the last 20 years. If I just play on any stock of the s&p 500, the result is superior to buy and hold. For example, running on spy, huge returns in 2008 and 2020, only one small loss year out of 20 years. But if I play on all s&p 500, the return/share is still very high, but sharpe comes down as holding one instrument. Anyway, the capacity of the strategy is lots of billions and 5-10 times higher return than buy and hold but at the same sharpe level. If I tried to improve the sharpe to 4 by enforcing mkt neutral, the capacity decreased to a very low level mainly because mkt tends to move collectively. The strategy does not hold position overnight. Anyone has interest in this strategy?
Is that 20yrs backtest data or 20yrs live running, presume the first, things started to change in 2018, and inverted the past year it doesn't quite work the way it used to which magnifies backtest uplifting of results. Actually it doesn't because that's not how it works, there are too many metrics so to keep it simple, what is the per monthly/yearly return profile for the past 20yrs, simple spreadsheet compared to actual S&P, I know funds who are interested in diversification strategies but the returns profile is what they look at.[/QUOTE][/QUOTE]
If your strategy accesses highs or lows of daily data, then it is suspect, especially when applied to equities. To confirm your strategy, you need to access intraday data to confirm that your entries would have been possible, and/or run it in real time for a period that's long enough to give you confidence that you have something worthwhile.
That is kind of beginner’s mistake assuming entry at high or low. My entry is at the vwap of the day.