Thursday, April 29, 1999
The P/E and Bond Yields: A Valuation Conundrum
Warren Buffett has said that equities are bonds in drag. With one, the yield is known, because it's usually fixed and it's printed on the instrument or in the prospectus. With the other, starting invested capital (face amount of the bond, to identify the bond analog), price, and current return on invested capital (the bond's face yield) are known. Thus, you know the current earnings yield (earnings divided by price) of the equity. What you don't know is what the coupon (earnings) stream will look like in the future. To figure that out, there are a heck of a lot of variables to consider, but the one that always gets the most play in figuring out what a company is worth is the earnings yield (or, inverting back, the P/E).
Single-point valuation analysis, i.e. the "P/E ratio" is about the least effective way to look at a company. The reason why the Rule Maker approach to valuation makes sense is because it eschews single-point valuation metrics. They're really not informative at all on their own, and when you have the data necessary to give enough context to this data point, then it becomes pretty much unimportant alongside more informative data points. Price/book, price/sales, price/forward P/E really mean little if you don't have data on margins, capital turns, cash conversion cycle, the plan for the company's capital buildout, and some idea of how much resources the company will need to grow its earnings by the ubiquitous "15% consensus EPS growth rate" that seems to be the equivalent of "I don't know" when you poll people on the company's forward plan. It's massively important to look at the income statement in relation to the balance sheet and cash flow statement in figuring out what today's P/E or earnings yield means.
The yield on invested capital is a constant condition for a bond. It's the face yield on the bond. A 6% $1,000 bond is going to pay $60 per year for the duration of the bond, assuming it doesn't get called or converted. Out to the maturity of a fixed coupon bond, that doesn't change. The only thing that's going to change the value of the bond, assuming its credit quality doesn't change and there are no weird call features, is prevailing interest rates. That changes the yield to maturity of an equity, too, because it changes the company's cost of capital and the discount rate used to deflate the future earnings stream to a present value, and it can also effect the amount of future cash flows the company can generate.
If you're paying 4.5 times invested capital and the company returns 14% on invested capital, then your current modified earnings yield (I'm calling it modified because it's not quite earnings divided by price because the common factors don't cancel each other out if you divide the one ratio by the other) is 14% divided by 4.5 = 3.1%.
If you pay par for a bond that has a face yield of 6%, then you're paying one times invested capital for an instrument yielding 6% on invested capital. Ignoring for purposes of simplicity the tax issues (which are, I caution, substantial), the Bore ratio on that would be 6% divided by 1 = 6%.
On the initial capital outlay -- the price the investor pays for that interest in the company -- the current yield can be low, just as if you're buying a 9% bond in a 6% interest rate environment. With equities, however, your return on incremental capital can be huge multiples of the return on capital that you're getting on your initial outlay. If you're investing in a company with a 40% return on capital and good growth prospects, you'll likely be paying around 10 times or more the amount of capital currently invested in the business. That would put your current modified earnings yield at 4%.
But the reinvested earnings, or coupon income in this example, is being reinvested in new capital projects (or new bonds, to round out the comparison), at par, that yield 40%. If you run that out over a long period of time, then you can pay a big multiple of earnings today and still show supernormal returns on your initial investment. Think of it this way: In an 8% interest rate environment, you would jump at the change to buy a special 5% bond that gave you the right to reinvest that coupon income at par in bonds of equal quality that paid a face rate of 40%.
One reason many value investors won't invest in the most dynamic enterprises today is due to their unwillingness to pay above certain arbitrary multiples to earnings or book value. The reason why the Rule Maker portfolio makes sense in many ways is because Tom Gardner and the guys that run it are very cheerfully optimistic about the future. They figure these globally dominant companies can keep on pouring on the capital and generating huge returns on capital. If you have a company growing invested capital at 25% per year for 30 years and returning 25% on invested capital, you can pay huge multiples to earnings or capital today and do very well over that holding period. Value investors by nature feel uncomfortable forecasting those growth rates, because they often don't happen. It is uncommon, but the whole point of the Rule Maker approach is to narrow your investment universe down to the companies that have the best chance of making that happen.
But it doesn't mean that sort of growth doesn't happen. It has and it does. The value investor wants a margin of safety, however. The goal of the value investor, though, shouldn't be to purposely underestimate what a company can realistically do, just as no "investor" should purposely overestimate what a company can do and then attach a big discount rate to that to compensate for over-enthusiasm.
The goal of investors should be to develop the best guess possible as to what a company can do going forward and then attach whatever hurdle rates they feel match their requirements. With a company like Wal-Mart, you would have been called insane by your colleagues if your 20-year model of the company in 1970 forecasted exactly what the company did do over the ensuing 20 years. But you could have discounted your cash flows at 20% per year and still have derived a huge multiple to 1970's market price as your intrinsic value.
So again, this points out that you can't really tell what the heck a company is about by just looking at the P/E. You've got to have some sort of return on equity, return on net assets, or return on invested capital information, as well as a forecast of a capital buildout plan, to really have an idea of how to value something. Just as in valuing a bond, you need to know how much capital is being used, the yield on that capital, and your own hurdle rate for discounting that to the present. Looking at just P/E without considering what the company can do in the future, specifically, is like looking at a bond's current yield without having any idea what interest rates are today, what they will be tomorrow, and what the coupon income from that bond will look like until maturity. It's impossible to value a bond that way, so I don't know why people try it with equities.
Call Your Boss a Fool.
|Recent Fool on the Hill Headlines|
|Fool on the Hill Archives »|