Hey all, I am not an experienced trader, however, whenever I am looking to do some backtesting or evaluating results, my mind automatically goes to "how would this have done compared to just buy and hold?" I have done some simple backtests and I automatically dismiss the system because they are worse than just buy and hold on a total return basis. However, maybe I am wrong with this mindset? My question is, am I correct in my way of thinking when evaluating trading results? Why or why not? Thanks!
Risk-Adjusted return is usually a better way to look at performance. https://www.investopedia.com/terms/r/riskadjustedreturn.asp My favorite one is the Ulcer Performance Index.
Being aware of buy and hold returns isn't nothing but I focus more on CAGR compared to maximum drawdown for my strategies. I understand why people find comfort in buy and hold but when the S&P has annual returns of roughly 10%, depending on timeframe, along with semi-regular drawdowns of 25%-35% with the occasional drop over 50%, it concerns me. I need anticipated drawdowns to be somewhere in the ballpark of annual return, at most double, not several multiple of it. A huge drawdown is when people get emotional and start thinking about selling everything to get a little catharsis which is the worst possible thing to do at that time and accounts for a lot of the underperformance active traders experience. Nobody knows the future and past performance doesn't predict future results yada yada but the equity strategies I gravitate to prioritize smooth return with superior risk adjusted performance. If the returns must be juiced to get over the buy and hold annual return hurdle, leverage is available. As long as risk adjusted expectations are superior, leverage shouldn't be shied away from. With that in place I look for strats that are non-correlated to equities and for me that equates to option selling, specifically iron condors, butterflies (if I'm feeling really precise about direction), and calendars (if volatility is really low since calendars are long vega unlike condors) depending on volatility and the specific underlying. Farming premium doesn't make up the bulk of my activity but it did outperform in e.g. environments like 2022 when stocks and bonds got equally rekt.
You cannot compare to B&H. You need to look risk adjusted. Best or most common is to use Sharpe Ratio as measurement, on how good the strategy is. If you want to beat the S&P500 or Nasdaq as B&H, you can just buy UPRO or TQQQ as built-in leveraged ETFs here. And you should outperform your benchmarks.
The only way I see you can make consistent returns which outperform your Buyandhold benchmark is daytrading based on Momentum plays, so you would care on stocks having their opening gap and then their common behavior thereafter. Then you can make by far more returns WITHOUT having more leverage than your Buyandhold comparison, if that is really important to you. But usually it is seen only risk adjusted, so the drawdowns play an important role. In that case you would need leverage when your absolute returns on your better risk adjusted strategy is smaller than your Buyandhold benchmark return. You also need to consider fees for paying interests on a non-zero interest rate environment, as the use of additional leverage cost you some extra money.
If you are considering daytrading stocks on Momentum plays, you have two options. You can either trade the Order Imbalance on the market opening, meaning that mean-revert trade but you need to be market neutral here. Or as second option you can also trade Momentum Breakout on opening gaps, maybe with entries on shorter term reversals to get better risk-rewards trades. I already posted in the past some successful youtuber here trading this second option, if you want to make a further search.
Hi Mike, ignore the replys above, they’re inexperienced traders and they replied nonsense. Here is the thing. Because you’re inexperienced, whatever you’ll do will initially fail, and if it will work, then it will be due to luck and won’t last long. We all went through it. Trading is a craft, you need to slowly chip away at it, and it takes years of frustration. You need to be realistic and stay committed. To your question, yes, the strategies need to outperform buy&hold, but you need to know which part of your system is letting you down, because sometimes a “bad” system can be turned into an OK system with some tweaks, but obviously it takes experience to know what to tweak. In your other thread, I wrote “computer tested methods will down the road result in inefficient use of trading capital vs learning to trade and manual backtests (you have to pick your poison)”. What I meant is that computer tested methods usually lack robustness (“one size fits all”), are dependent on market conditions, have large drawdowns and the annual profits are very low. On the other hand, (all things being equal) rule based discretionary trading (pattern recognition) is harder to master, but less drawdowns, and much higher returns, therefore more capital efficient. Obviously, most people fail with both methods, but either way, you’ll need to decide which way you want to go. Hybrid, i.e. mixture of both works well with stocks, but not as good with other markets. You’ll need to experiment to find out which approach fits your own personality. Meanwhile you can do manual backtests of say 20 trade samples, and it should start giving you some feedback. If that seems to work, then you’ll need to start to fine tune it (not optimising) and do far more through research. It takes commitment and time. You can post your manual backtests here (charts with the trading logic) and the spreadsheet with the results. That way you might get more accurate reply to your question. Manually backtesting 20 setups and marking up the charts takes very little time, few hours max.
That’s not true at all, there are so many ways to outperform the market. Also, advising beginner to day trade is inappropriate advice, dyatrading is far more demanding than EOD trading. PS: the other poster deleted his reply after he saw my response. Interesting. Come on guys, we all were beginners once, so why not help those who are struggling when they politely ask for help, what's the point of trying to "impress" them with the nonsense with options, etc, that the other poster (not you @TrAndy2022) deleted.
20 trade samples seems like a way too small sample size to get a true sample of if the pattern truly has an edge though, no?