In addition, I believe a perfect storm for earnings has formed - falling profit margins, debt destruction, deflation and a credit crisis. This applies to the Wall Street strategists and their plummeting “top-down” or index estimates. Many of these strategists started with S&P 500 top-down earnings estimates north of $100 per share for 2008-2009. With the S&P at 1200-1300, we were told that stocks were cheap relative to risk-less Treasuries.

I stated many months ago that I felt stocks were still in a bubble (The Fed Model, June 9th) and that there were 2 problems with the $100-per-share estimate. First, I believed the $100 estimate was way too high, and second, Treasuries were the wrong asset to compare equities to. Arbitrarily picking the 10-year Treasury is as useful as picking the price of tea in China - to me, anyway. It’s also like comparing other risk instruments -- namely corporate bonds -- to Treasuries.

Instead, we compared them to riskless Ginnie Mae’s with yields north of 6% (some “specified pools” can be had with yields close to 7%). While I no longer think stocks are in a bubble, I still don’t find them cheap compared to other investment options. When stocks yield 3% -- and I’m not convinced that earnings estimates are correct, nor will they grow -- I still find the “equity-like” returns of Ginnie Mae’s attractive.

For example, IBM (IBM) recently came to market with a multi-billion-dollar corporate bond deal that yielded approximately 400 basis points above comparable treasuries (never mind that IBM has historically been a very savvy issuer, issuing just as spreads widen over Treasuries).

I suppose if you live in a vacuum and consider the fact that IBM paper does look cheap on a historical basis vis-à-vis Treasuries, they don’t look cheap versus risk-less (from a credit perspective) Ginnie Mae’s and soon-to-be-explicitly-backed Fannie (FNM) and Freddie (FRE) paper.

Flawed Thinking

Let’s say you actually believe Wall Street’s earnings estimates. Further, let’s assume you have no idea how to buy Ginnie Mae, Fannie Mae or Freddie Mac paper - or, for that matter, corporate bonds. All you have to choose from is 10-year Treasuries and stocks. Since no one brings up the Fed Model much anymore, allow me to do so here and now.

The Fed Model was developed to determine if stocks were “cheap” relative to 10-year Treasury Notes. All you had to do was take the inverse of the yield of the 10-year Treasury note (the Price/Earnings ratio of the 10-year) and compare it to the P/E ratio of the S&P 500. When the P/E of the S&P was lower than the P/E of the 10-year, stocks were deemed to be “cheap.”

This flawed thinking worked very well from 1982-2000, a period of disinflation and the strongest secular bull market for stocks on record. The problem, of course, is that this was the only time this concept worked. If we go backwards in time to the 1950’s and stocks traded at 7 times earnings and Treasury yields were 3%, stocks would have been undervalued by 75%. Furthermore, if we look at Japan’s 1% rates, stocks would be cheap at anything under 100 times earnings.

In the chart of the Fed Model below, courtesy of Ned Davis Research, stocks have looked cheap for the past 6 years, a period when many have faced unprecedented volatility and made no money. What’s most disconcerting is that this doesn’t take into consideration how much money has been lost in the real estate market and the credit market.

Sub-prime mortgage issues, now spreading to commercial real estate, may end up dwarfing losses in equities. We’re just now beginning to see the issues in the Alt-A and Option-ARM space. In other words, get ready for more pain, as if we haven’t endured enough already.


Click to enlarge

Worse yet, if you had listened to Wall Street’s top-down estimates a year ago (yes, the same bunch who completely missed the largest unwinding of credit and risk in history), stocks would have seemed cheap at 1500 with estimated earnings of $100 per share. The math goes like this:

1500/100=estimated P/E of 15.

1/10 year Treasury note Yield (4%) =25.

We then apply the 25 P/E of the 10-year to find the “fair value” of the S&P of 2500. Flawed thinking at its best, I would say. And not only is it flawed, it fits the sell-side motive of selling all of this stuff to investors.