Surprise!

Often skewness is elbowed aside while Sharpe ratios are proffered to the Gods.

Reminds me of jumping on the subway in Beijing.

The S&P 1500 Momentum index has negative skew - more than you see with the S&P 500 - which accounts for higher returns, exactly the reason why value often beats growth.

The Momentum++ strategy has very positive skew.

Positive skew is a bit like getting a Kinder Egg for the first time.

In this case, skinnier left tails (the chocolate) with a shot at hitting the jackpot in the right tail (the toy inside!).

The next chart is what happens to both the S&P 1500 Momentum Tilt and Momentum++ when the S&P 500 has positive returns.

Tilt's beta is 89%, pretty low.

Whereas ++'s is 104%, which means, on average ++ will be much more profitable than the S&P 500 during upswings.

Now for the magic.

When the S&P 500 drops, Tilt's beta is 97%, 8% more than during upswings!

Classic momentum strategies like Tilt greedily grab at losses while letting profits escape.

Negative convexity risk is why momentum returns more over time

Exactly the same reason for the growth/value premium - but bigger - the only reason why classic momentum has higher returns over time is because of non linear risk which is too often glossed over.

On the other hand, Momentum++ has a lower beta of 69% during downswings.

I need more data to be sure of ++'s results, but nevertheless it's a good example to show why skewness and convexity matters.

Also calculating Big-O Sharpe efficiently with the Lazy Backtest IDE will help give me more reassurance with these figures and is top of the to-do list.

In any case, it's becoming increasing clear that classic momentum is a bit like a Kinder Egg with a nasty surprise inside - ready to blow up whenever volatility spikes.