I saw a chart on stocktwits with high beta divided by low beta stocks (performance wise).
It correlates with my 5 Stock portfolios, e.g. when high beta stocks (“risk is on”)
are doing well, my 5 Stock portfolios jump up, otherwise they underperform.
Any Idea to build a market indicator here on p123 in this direction, so we can
backtest this idea?
Best Regards
Andreas
This is a great idea! You can do this with two custom series: one for high beta stocks, one for low beta stocks.
Steve
Andreas,
Great concept, and in theory, based on CAPM model, high beta stocks should yield greater returns. Unfortunately, in practice, this is not the case.
I actually did a back test using Beta3y ranking system, the performance will indicate there is no trend to support the CAPM model.
-Luis
When Hi-Beta stocks perform better than Lo-Beta stocks the model signals investment in the stock market, otherwise investment in bonds.
Andreas Himmelreich noticed that it correlated well with his 5-stock portfolios, e.g. when high beta stocks were doing well, his portfolios went up in value, and otherwise it under-performed. Subsequently Steve Auger of Stock Market Student provided an algorithm to model this.
Hi-Beta and Lo-Beta are defined as the Market-Cap-weighted average of the performance over a specific period of the S&P 500 stocks that meet the criteria “Beta > 1.5” or “Beta < 0.75”, respectively.
Hi-Beta and Lo-Beta are smoothed with a 40-week Exponential Moving Average (EMA), and the signal to enter the stock market is given when the 40-week EMA of Hi-Beta > 40-week EMA of Lo-Beta.
It clearly signaled the exit from the stock market before the two major downturns, in 2000 and 2007, and also the re-entry early in 2003 and 2009. The model would have avoided losses of 40% and 45%, respectively, for the two major down-markets. Figure-3 and Figure-4 below.
So a P123 series “Hi/Lo Beta Ratio” would look something like this?:
(UnivSum(“beta>1.5”,“Pr4W%Chg”)/Univcnt(“beta>1.5”)) / (UnivSum(“beta<0.75”,“Pr4W%Chg”)/Univcnt(“beta<0.75”))
Then test like:
ema(40,0,getseries(“Hi/Lo Beta Ratio”))>1
