Firstly even if you know with certainty the “Fair value” of an equity or equity index or for that matter any asset or any relationship between assets the path to convergence to FV is rocky. I spent decades in the RV space before learning macro and have …
Traded many absolute and statistical arbitrage relationships. Convergence to FV for many of these strategies often depended on macro conditions. This generated track records that often were simply levered beta with bad drawdowns and expensive transaction costs and fees.
The path to convergence generated p/l volatility which impacted risk-adjusted returns. The returns from knowing FV just don’t compensate investors for the risk significantly more than owning a passive beta portfolio. Tl;dr alpha is hard to get.
So why try at all? TBH it is arrogant to believe you or I have alpha and it is unlikely that we do. But that’s another debate. The goal of this thread is to review equity valuation methods. There is no right answer and all can be used if used appropriately to triangulate FV
Let’s start with DCF. A fixed-income instrument has a price. With that price, we can precisely determine its yield to maturity. As a side note this YTM is generally accepted but in actuality ignores the scientific approach of discounting each flow at its appropriate discount rate
But I digress and won’t explore that rabbit hole. The good news is fixed income securities have known cash flows. But even then the price or yield may not be fair. Let’s for now assume it is. The FV is the yield is the price. No degrees of freedom
Equity has no fixed payments. Even dividends change over time. There is no maturity and no earnings or cash flow certainty. That creates a massive assumption necessary to make on cash flows. Also, equities are risky in a way that government bonds are not. For that reason…
The Appropriate rate for discounting future flows is not the risk-free rate but an unobservable rate that most model with the risk-free rate plus the unobservable risk premium. So now we have two assumptions. Future flows and discount rate. Let me add another.
Dividends if companies even pay them are only part of the annual flow. For companies that earn money the company’s value increases by the amount they “earn” over the dividend they pay. Most people consider that a flow worth of discounting. I do too. But here’s the rub. GAAP lies
Accounting earnings include a variety of conventions that make it difficult to compare the Net income of two companies. The amount of leverage, the number of depreciable capital goods they own, and the taxes they pay all differ even if within sectors. In order to adjust that
EBITDA is commonly used. Earnings before interest, taxes depreciation, and amortization is the acronym. This is a good place to start. But it itself can create a bias because it can favor or disfavor valuable decisions regarding capital investment, leverage, or tax jurisdiction
Nothing’s perfect. EBITDA is the closest. The last step is how one handles the maturity of the investment. No one knows how to value a stock today and so ascribing a terminal value for the stock 30, 50, or a century from now is a fool’s errand. The way this is handled is typically
By assuming a company is a perpetuity and counting on its long duration for the terminal value to be a small part of the FV. But again it’s a bit assumption.
So a DCF has major assumptions
1. What is the discount rate
2. What are the growth assumptions for the business
3. What accounting measure should be used to determine the cash flows
4. What is the terminal value assumption.
Do you see why this can’t be used for alpha?
Why bother with projections? Let’s just look at static ratios! Everyone uses P/E why don’t you Andy? Well, I do! In fact, P/E is a reasonable shortcut for assumption guessing However it has a fatal flaw. It completely ignores the discount rate. Regardless of your current view
On P/E’s I hope you accept that when interest rates fall companies are more valuable all else being equal assuming the earnings don’t fall. So today P/E’s are high. Part of that is Rf interest rates are low. I know what some of you Europeans and Japanese are saying. Our interest
Rates are even lower and yet our P/E’s are by and large lower. This is because your Risk Premiums may be significantly distorted. It’s for another thread but the big takeaway is Japanese and European assets and their economy suck because financial conditions are not loose enough
Ok, so what is my favorite heuristic? I care about the index most of all. I make conservative assumptions regarding projections of businesses and I want interest rates in my thinking. I also believe equities are long-duration assets that respond most to long-term risk-free rates
My basic model is to use the simple inverse of P/E called Earnings Yield or EY and reduce that by the risk-free 30 yr rate. What comes out is an approximation of the risk premium. I then track that through the long history of US stocks and see where it falls. Currently that …
The number is 2.75%. Pre Covid It hung out around 2.25% And so I sorta like stocks based on this AGAIN not alpha signal. Some context during the dot com bubble this number was negative. Meaning based on earnings risk-free treasuries were a better investment. Similarly, 2007 showed
A lower number. That is my view and again I use many models to triangulate my value indicator. I then discount heavily the predictive power of that number and add it to other value and non-valuation-related indicators to generate a signal. I then humbly ask myself
Do I possibly have any alpha? This is a difficult game I have been playing for 35 years and I still don’t know the answer but lean to the assumption that I do not.