I read some very interesting concepts about this in a book called Skip the Dip. The idea is to have a number of complementary simple auto-trading strategies. For instance, you can have:
long only momentum trend following strategy with a basket of stocks that tend to move a lot, and exits when it loses momentum. This tends to have low win rate but very good expectancy because they capture big moves. It loses on choppy markets and skips downturns. When a couple of names rip, it covers the tiny losses of the others that are chopping. Works well with shares.
mean reversion strategy with very selective criteria on a basket of stocks. This one has a high win rate but makes small moves. Pads the equity curve. This tends to work best with credit put spreads but can be done with stocks.
long volatility strategy, which loses most of the time in a tiny drift and when VIX gets crazy, it makes outsized gains when the other two above are not trading as the entry conditions are not being met. This can be done with VXX or UVXY shares.
You could add here some broken wing butterflies for income when certain technical criteria are met, to pad the equity curve a bit more.
After very promising backtesting results, I am now setting up the algos in paper trading, and then move into trading 1 share per ticker on signal. I am curious as to any insights from @tod