Lately, I've been looking into Macro Factors. As a starting point, I have tried to come up with factors based on the high yield spread and the 1 year US treasury yield.
My idea was that when the high yield spread is high and rising, I would want to invest more in stocks that have done well in those enviroments and hence have a positive beta to changes in the high yield spread.
The same for when the high yield spread is high and falling. I would then want to invest more in stocks that have low (negative) beta to changes in the high yield spread.
I noticed that when I take a lookback that is too far, the linear regression outsputs will more often become NA. More importantly, I found that while in the US these factors worked decently, I didn't get the same output in Europe.
See below for the ranking output for the Easy to Trade Universe.
I have tried something similar, but it mainly creates higher turnover for me.
Something I really would like to have is macro as base to create dynamic weight for the factors.
Like the higher volatility in SPY, the less Beta weight. With use of a volatility limit condition to create a true or false condition, it would create very high turnover at times. If it would be dynamic, it think that would be less of a problem.
I have a few macro flags that I use just to get a feel of the economy. I generally put these in functions and then add it up for a total score among some other factors. In one of my accounts I will adjust my cash positioning up or down depending on the score, or I will put some money into a tail risk ETF with out of the money puts (like $CAOS) for a hedging
Total Inventories
eval(CLOSE(0,##INVTOT) > EMA(25,0,##INVTOT),0,1)
Retail Sales (Intuitively think this is a good leading indicator. If people see dark clouds on the horizon at their job they curb spending)
