04-03-2014, 09:13 AM
(This post was last modified: 04-03-2014, 09:14 AM by AlphaQuant.)
Rather than looking at a time series to derive correlation (which is just a number), one is better served going back to first principles.
Simplifying valuation of an asset:
price is proportional to (earnings* growth - debt*int rate)/(riskfree rate+asset risk premium)
as rates rise, debt*int rate is -ve to price
risk free rate rises is -ve to price
let's call asset risk premium a scratch
the crux of the matter then is whether earnings growth is sufficient to overcome the 2 -ve factors when rate rises, and that differs for each counter. This is different from bonds which by virtue of being a fixed income instrument (hence no chance of earnings growth), has to be -vely correlated to interest rates.
So i don't necessary think that REITs are -vely correlated with rates - the essence is to find something where earnings growth > debt*rate rise+ discount factor from rising riskfree rate.
It's rather tough, since it's 2 -ve factors against one, but it's not impossible.
Simplifying valuation of an asset:
price is proportional to (earnings* growth - debt*int rate)/(riskfree rate+asset risk premium)
as rates rise, debt*int rate is -ve to price
risk free rate rises is -ve to price
let's call asset risk premium a scratch
the crux of the matter then is whether earnings growth is sufficient to overcome the 2 -ve factors when rate rises, and that differs for each counter. This is different from bonds which by virtue of being a fixed income instrument (hence no chance of earnings growth), has to be -vely correlated to interest rates.
So i don't necessary think that REITs are -vely correlated with rates - the essence is to find something where earnings growth > debt*rate rise+ discount factor from rising riskfree rate.
It's rather tough, since it's 2 -ve factors against one, but it's not impossible.