In recent weeks, there has been considerable discussion among "pundits" about the impact of interest rates on the stock market. It has been asserted that the increase in rates since the Spring of this year indicates that stock prices will decline. It is important to review stock market metrics which utilize interest rates as an input in order to better understand just where we are and where we are likely to go from here.
There are a variety of metrics or "rules of thumb" which strategists use to gauge the market and its likely direction. Some of these metrics take no account of prevailing interest rates whatsoever. Thus, the Cyclically Adjusted Price Earnings (CAPE) ration looks at the current price of the market and compares it with the average earnings over the past ten years to derive a ratio which is then compared with similar ratios in the past to determine whether it appears "high" or "low." This metric will give you the same answer when the federal funds rate is 15% that it will give you when the federal funds rate is 0%. Similarly, the Tobin Q looks at a ratio between two numbers calculated by the federal government and uses that ratio to determine whether the market is overpriced. As with the CAPE ratio, the Tobin Q metric does not take account of current interest rates. Analysts who have followed these metrics slavishly have already missed one of the major rallies in stock market history. I have suggested in previous articles here and here that interest rates are a necessary component in any strategic assessment of the market.
Before going further it is important to understand why interest rates are relevant to equity valuations. There are at least three important reasons. First and most fundamentally, equity valuations ultimately depend upon the discounted present value of future cash flow; the discount rate used in this calculation should depend - at least in part - upon current interest rates. A low discount rate will produce a higher present value and, thus, low rates tend to produce higher equity valuations. Stated more simply, in a world in which it is hard to get a 3% return on one's money, a piece of paper which has a dividend yield of 3% and an earnings yield of 6% tends to be worth more than that same piece of paper is worth in a world in which it is easy to get a 12% yield on one's money.
Secondly, investors will have a tendency to demand a lower dividend and earnings yield from stocks when those investors do not have attractive alternatives in the form of high yielding bonds. While there are all sorts of caveats, there is definitely a tendency for low interest rates to push or, perhaps, coax investors in the direction of equities with a resulting upward pressure on prices.
Finally, corporations themselves will engage in "financial engineering" if cheap debt is available. This is sometimes described as "balance sheet optimization" and takes the form of borrowing money to buy back stock, engaging in cash for stock acquisitions and participating in leveraged buy outs. In each case, returns on equity can be enhanced by trading low yielding debt for equity with a higher earnings yield. As long as the after tax interest rate on debt used to fund repurchases is lower than the earnings yield (the inverse of the price earnings ratio) of the stock being repurchased, the net effect of borrowing to buy back stock is to increase earnings per share.
If it makes sense that low interest rates are good for stocks, then shouldn't it be the case that an increase in interest rates will result in a decline in stock averages? Not necessarily. In the current market, a strong case can be made that an expectation of higher interest rates is already "baked in" to the pricing of the market.
I am aware of at least two market metrics utilizing interest rates to assess the market. Let's take a look at what they tell us.
The Fed Model. The first metric is the "Fed Model" which utilizes the 10 year Treasury Bond yield and compares it with the earnings yield on stocks. It was popularized in the 1990's (I believe, by Ed Yardeni) and has the virtue of simplicity. The Table below provides a series of 10 year Treasury Bond yields and corresponding price earnings (PE) ratios and projected S&P 500 index prices. I am using S&P data from Barron's (earnings of $87.20, a price of $1709.91 and a PE of 19.50). I have used the low Treasury rate from last summer, the current rate of 2.73%, and some potential future rates.
|S&P 500 Index||$6218.64||$3210.19||$2920.74||$2508.50||$2192.75||$1947.16||$1754.20|
Under the Fed Model, the S&P 500 Index is actually priced for a 10 year Treasury Bond yield of over 5%, well above the current level of 2.73%. When the low rates were reached last summer, stocks were priced at levels building in enormous expectations of rising rates, which expectations have not yet been realized. At the current rate of 2.73% the yield on ten year Treasuries implies a PE of 36.6 and a pricing of the Index at almost twice its current levels. So - yes - higher rates have reduced the price of the Index as suggested by the Fed Model, but the reduction has been from a level nearly four times the current price down to a level nearly twice the current price! And rates could go up a long way before they would begin to reach levels which would suggest that the market is overpriced.
The Timmer Model. The second metric which uses interest rates is the Timmer Model and is based on insights in an excellent article by Jan Timmer . Focusing on the phenomenon of financial engineering, this model is based upon the after tax interest rate paid by corporate borrowers. The reasoning is that - as long as this rate is lower than the earnings yield on common stock - corporate management will be able to increase earnings per share by borrowing money and buying back stock. This will tend to lead the earnings yield to equal the after tax borrowing cost. Using this metric, one can estimate the PE that would be associated with a given interest rate and use this PE to calculate a target price for the Index.
It is, of course, impossible to develop a universal corporate borrowing rate. I have conservatively used six per cent as a rate at which many of the large corporations can borrow money for 10 years. Readers should be aware that many, many large corporations can borrow at much, much lower rates. For example, Apple (AAPL) has 10 year bonds priced to yield 3.64%, Verizon (VZ) 3.55%, Bank of America (BAC) 4.06%, and AT&T (T) 4.15%. In this regard, the S&P 500 Index is weighted by market cap and so the impact of large companies with low borrowing costs - such as AAPL, Exxon (XOM), Microsoft (MSFT) and Verizon - is very large. Thus, the 6 per cent borrowing cost I am using for companies in the Index is very, very conservative. At any rate, I have assumed a 35% corporate tax rate and I have calculated the after tax borrowing cost, the PE, and the Index price which would correspond to various corporate interest rates using this Model.
|After Tax Cost||3.90%||4.55%||5.20%|
Under this Model and using the 6 per cent rate(which is almost certainly considerably too high) , the market is currently priced nearly 20% below target. The current pricing would be warranted only as the interest rate faced by large corporations began to near 8%. Once again, we have a long way to go before interest rates will start flashing a "Sell!" signal.
I do not believe any model should be accorded a talismanic reverence and I make no exception for the Fed Model or the Timmer Model. However, I believe it is also a serious mistake to worship at the altar of a model which ignores interest rates. I also believe that, as long as the earnings yield on a stock is substantially below the company's after tax borrowing costs, there will be a strong incentive for financial engineering through share repurchases and, to a lesser extent, cash for stock acquisitions. These forces will lead to financial engineering which will provide buying pressure in the market and will gradually remove shares from the market. While these forces may not actually force the market to the exact levels suggested by the Model, they will exert a powerful force and this force will tend to become more powerful to the extent that the market deviates further from the level suggested by the Model because the incentives favoring financial engineering will be enhanced by such deviation.
I think that the current environment continues to favor companies with strong balance sheets that can take advantage of generous credit markets and, in this connection, I am still bullish on AAPL, MSFT, Cisco (CSCO), Seagate (STX), Western Digital (WDC) and Johnson & Johnson (JNJ). All things being equal, higher interest rates are bad for stocks. But the market has already priced in substantial increases in interest rates so that interest rate increases do not necessarily constitute sell signals.