A Mean Reversion Strategy with 2.11 Sharpe

[quantitativo]

Hey guys,

Just backtested an interesting mean reversion strategy, which achieved 2.11 Sharpe, 13.0% annualized returns over 25 years of backtest (vs. 9.2% Buy&Hold), and a maximum drawdown of 20.3% (vs. 83% B&H). In 414 trades, the strategy yielded 0.79% return/trade on average, with a win rate of 69% and a profit factor of 1.98.

The results are here:

Equity and drawdown curves for the strategy with original rules applied to QQQ with a dynamic stop​

Summary of the backtest statistics

Summary of the backtest trades​

The original rules:

• Compute the rolling mean of High minus Low over the last 25 days;
  
  
• Compute the IBS indicator: (Close - Low) / (High - Low);
  
  
• Compute a lower band as the rolling High over the last 10 days minus 2.5 x the rolling mean of High mins Low (first bullet);
  
  
• Go long whenever SPY closes under the lower band (3rd bullet), and IBS is lower than 0.3;
  
  
• Close the trade whenever the SPY close is higher than yesterday's high.
  

The logic behind this trading strategy is that the market tends to bounce back once it drops too low from its recent highs.

The results shown above are from an improved strategy: a better exit rule with dynamic stop losses. I created a full write-up with all its details here.

I'd love to hear what you guys think. Cheers!

 

[hilmy83]

Nice, I like a nice description and numbers to back it up. You live traded this yet?

 

[quantitativo]

Nice, I like a nice description and numbers to back it up. You live traded this yet?

Thanks! Not yet, I'm forward-testing it first. Also, I don't think this strategy is ready to be traded as is, because the exposure time is too low (only 15%). I want to change it to at least 30% before trading it. So far, I'm thinking about 2 ways to achieve it:
1. Add complementary strategies to it. If I manage to add 2-3 strategies with similar characteristics (i.e. similar distribution of trades) such that I can reach 30% of time invested, that would lead to +25% in annual returns;
2. Change this strategy such that it trades the individual components of the Nasdaq 100 simultaneously, in parallel.
Let's see how it goes. I'll post what I learn about these avenues here as soon as I can..

 

[Businessman]

414 trades.

Sample may not be large enough.

 

[schizo]

I'm not sure how to accept this. My mama used to say, when it looks too good it ain't what you think. Sorry, just kvetching.

 

[murray t turtle]

%%
MAY help;
+SPY average is closer to 12 %, depending on start time .
But just too add some '' food for thought '' as Schizo says ;
SARK, ETF up >18% YTD/ but SPY beats SARK a lot, 12 months+ longer .......
But another advatage of SH or SARK, could keep you from wrongly messing witha good investment, ETF invesrtment LOL .
Some times I beat my investmenst daytrading + sometimes not .
And diversification in time could help anything.
Win rate maybe a bit hi/ but I dont object to that number LOL

 

[quantitativo]

414 trades.

Sample may not be large enough.

Yeah, you are right. I'll test the 2 avenues to increase the exposure time (add complementary strategies, and trade Nasdaq 100 instruments simultaneously in parallel), and they should significantly increase the # of trades. I'll share as soon as I find more time

 

[quantitativo]

%%
MAY help;
+SPY average is closer to 12 %, depending on start time .
But just too add some '' food for thought '' as Schizo says ;
SARK, ETF up >18% YTD/ but SPY beats SARK a lot, 12 months+ longer .......
But another advatage of SH or SARK, could keep you from wrongly messing witha good investment, ETF invesrtment LOL .
Some times I beat my investmenst daytrading + sometimes not .
And diversification in time could help anything.
Win rate maybe a bit hi/ but I dont object to that number LOL

You are right! In fact, I tested a Long&Short portfolio using QQQ and PSQ. The results are in the full write-up, here's the link.

 

[SimpleMeLike]

Hey guys,

Just backtested an interesting mean reversion strategy, which achieved 2.11 Sharpe, 13.0% annualized returns over 25 years of backtest (vs. 9.2% Buy&Hold), and a maximum drawdown of 20.3% (vs. 83% B&H). In 414 trades, the strategy yielded 0.79% return/trade on average, with a win rate of 69% and a profit factor of 1.98.

The results are here:

Equity and drawdown curves for the strategy with original rules applied to QQQ with a dynamic stop​

Summary of the backtest statistics

Summary of the backtest trades​

The original rules:

• Compute the rolling mean of High minus Low over the last 25 days;
  
  
• Compute the IBS indicator: (Close - Low) / (High - Low);
  
  
• Compute a lower band as the rolling High over the last 10 days minus 2.5 x the rolling mean of High mins Low (first bullet);
  
  
• Go long whenever SPY closes under the lower band (3rd bullet), and IBS is lower than 0.3;
  
  
• Close the trade whenever the SPY close is higher than yesterday's high.

The logic behind this trading strategy is that the market tends to bounce back once it drops too low from its recent highs.

The results shown above are from an improved strategy: a better exit rule with dynamic stop losses. I created a full write-up with all its details here.

I'd love to hear what you guys think. Cheers!

Good Morning quantitativo,

Good looking curve.

If the money return for 25 years meets your personal trading goals, I would say trade it!

 

[semperfrosty]

I'm not sure how to accept this. My mama used to say, when it looks too good it ain't what you think. Sorry, just kvetching.

View attachment 340661

Yep.Words to live by.