Looking for a stear in the right direction on a relative value tool i am wanting to create. Essentially i want to be able to compare the bond yield/price with the equity price of a company in order to have a relative valuation so i can see company x’s share price looks rich vs the bonds or company y’s bonds look cheap vs the equity price. How should i go about thinking about this ? Ive seen some rolls that use dividend yield vs bond yield but that seems way to simple and factors out many differences between the two asset classes i.e. capital growth should outstrip cash flow in equities while in bonds its all cash flow and less capital appreciation. How do people think i should go about building such a tooland compare expected equity returns with expected bond returns ?
The challenge is that bond yields should be priced lower than stock yields to reflect the risk premium on stocks (and their increased volatility), so you are only going to find that stock yields are worse than bond yields when the stocks are massively overpriced. One way you might try to get around this it is to compare z-scores on historical yields, so that each yield is measured relative to its historic averages.
There’s a whole field of capital structure arbitrage that is a hedge fund strategy designed to see whether the current capital structure is equity-heavy or bond-heavy and likely to revert to some standard. I was never fully able to figure out how that works, but it’s a literature you can look for that probalby will help with your task.
I seem to recall that one approach is built on the merton model, saying that the stock price is basically the value of a call option on the assets, with a strike price equal to the value of total debt. From there, you try to figure out based on the assets and debt levels what the price of the stock should be, calculate what that means for the market cap, then figure out what that means the ideal proportion of stock to debt is. If it’s stock heavy, sell stock, buy debt, if it’s bond heavy, sell bonds and buy stock. How to get the hedge ratios right seems puzzling to me, though.
Great, thanks for that, some useful stuff for me to research. How about using a dividend discount model vs bond yield and using a z-score on that, could that be a good way of looking at this ?
If you are trying to get a yield rather than a price, I’m not sure a DDM is really going to take you anywhere. You can perhaps come up with a justifiable PE ratio that way, but if the stock isn’t actually at the price implied by the DDM, I’m not sure how you would compare it to a bond yield fairly.
I suppose you could take the ratio of stock yield to bond ytm and then look at the z-score the ratio compared to its historical distribution. You would definitely want the stock yield in the numerator, because it can go negative more easily and that will wreak havoc on your distribution. You probably would also want to scale up the bond yields to have similar dollar duration as the stock’s duration (the S&P index typically has a duration of 30 to 50), whereas the bonds might have 5-8 year durations. I guess that is how you might try to construct the portfolio (along with the volatilities).