Using Adjusted Close vs OHLC

[cgdutoit]

Hi -

It seems most EOD data services will provide OHLC and Adjusted Close.

If I'm correct, one should use adjusted close to calculate your return otherwise you'll get misleading results, in that backtesting.py would not magically account for stock splits.

Given that, this is easy for indicators that only rely on a close price as I can use adjusted close instead.

However, what do you do for indicators that need a combination of OHLC (Eg. candle patterns, many momentum indicators etc)? Given stock data services only provide an adjusted close, I can't mix OHL with adjusted close (apples and oranges). But likewise if I just use OHLC I'll get incorrect results as Close will not be accurate given stock splits.

Appreciate any insights,