podcast link including transcript: https://investresolve.com/podcasts/the-two-fundamental-drivers-that-determine-all-economic-regimes-podcast/
Asset classes perform predictably during regimes but regimes are difficult to forecast and time
- The desired diversification is actually diversification across real underlying economic factors. Different assets perform differently depending on economic circumstances. When we zoom out and look at quarterly or yearly data, we see the behavior of various asset classes in response to economic conditions is durable, even though on a daily basis there is noise or short term correlations can increase.
The 2 axes create the Growth/Inflation 2x2 Matrix
- Inflationary growth regime
- Deflationary growth regime
- Deflationary bust
How to combine assets to construct a portfolio
- Balance the risk
- If a portfolio is 50% in stocks and 50% in short term treasuries, 90% of the return is going to be explained by the stocks. [I think of this is as the portfolio’s beta or delta to each component).
- Maximize diversification
- Diversification ratio = Component volatility / Portfolio volatility
- The diversification ratio is an estimate of how many unique bets you have. So if you your portfolio vol is 1/2 you component vol you have 2 discrete bets.
- In practice, rather than equal weighting the vol contribution of the components, a manager may solve for the weights that maximize the diversification ratio. [I’d be interested in seeing how far the prescribed weightings would differ from equal vol weighting]
(remember the component or asset class volatility weights should be sized to be equal, so each asset is contributing the same amount of volatility. In practice, you can compute the weighted average component vol…option traders will see that if we square the inverse of the diversification ratio we compute the avg cross-correlation between the components!)
How the number of unique bets varies
- The quantity of unique bets varies based on component assets:
- Since the vols and correlations change through time (the average long term average correlations masks how much these correlations bounce around…stocks/bond being an extremely obvious example where there are positive and negative correlation regimes), the active management of such strategies jiggers weights in accordance to contemporaneous conditions (ie shorter lookbacks than long term-averages)