Cheapest Stocks in Two Decades based on Flawed Fed Model

There is a fascinatingly bullish article on Bloomberg yesterday, titled Cheapest Stocks in Two Decades Signal Bull Market. Its all about the so-called Fed Model, and what it means for stocks.

A quick excerpt gives the full flavor of the article:

"BlackRock Inc., Fisher Investments Inc. and Schroders Plc,
which manage about $1.4 trillion, say stocks are inexpensive
relative to bonds. Profit of companies in the Standard & Poor’s
500 Index, the benchmark for American equity, is growing faster
than shares, and represents a yield of 6.53 percent compared
with 4.65 percent for 10-year U.S. Treasury notes.         

The gap — the widest since 1986, according to data
compiled by Bloomberg — is encouraging investors because
earnings forecasts indicate the U.S. will keep growing, while
bond yields show confidence that inflation will stay in check."

There is a few problems with this sort of analysis: The biggest problem with the so-called Fed model is that its built on two assumptions: 1) That profits will stay high, despite being a cyclical peak and decellerating; and 2) that interest rates will stay low. If either of these variables move off their present readings by a significant amount, cheap stocks suddenly look a whole lot less cheap. (my view on valuation is that, based on the S&P500 earnings, stocks are neither cheap nor expensive. Regardless, bigger stocks are cheaper than smaller stocks).

The second issue is that Fed model double counts low interest rates. How? When rates are low, the cost of borrowing drops, allowing companies to finance cheaply, retire debt etc. Those savings and gains show up in earnings. That’s certainly what has been going on the past few years. If the next next step in your analysis is to then compare those earnings gains and valuation to Treasury yields, you are simply counting the impact of low rates two times.

The last problem with the so-called Fed model is that  I’ve never seen a proof that this is determinative or predictive of future market performance. When stocks became expensive via the model in the mid to late 1990s, they managed to rally another 3 years before finally rolling over.

If anything, the Fed Model teaches us that Valuation is a rather imprecise timing tool . . .


UPDATE: April 3, 2007 9:22am

As we have previously noted many, many times, P/E ratios go through periods of cyclical expansion and contraction. The expansion of P/Es during the 1982-2000 bull market was responsible for 75% of market gains; The present contractionary P/E cycle is partly responsible for the lowering the P/E of most stocks.

Miller Tabak’s Peter Boockvar reminds us that according to the Fed Model, in 1981 stocks with depressed earnings and sky
high interest rates were crazy expensive — just as the greatest bull market in US history began.


Cheapest Stocks in Two Decades Signal Bull Market
Michael Tsang, Daniel Hauck and Nick Baker
Bloomberg, April 2 2007

The Fed Model: Fix It Before You Use It
Jonathan Clements
WSJ, May 1, 2005,,SB111491292409921442,00.html

Print Friendly, PDF & Email

What's been said:

Discussions found on the web:
  1. V L commented on Apr 3

    In addition, they used deceptively optimistic “operating earnings” or excluding expenses pro forma earnings; using real GAAP S&P P/E’s the stocks are not that cheap. Moreover, they ignored the risk premium associated with stocks (bonds and stocks are not equal when it comes to risk; nevertheless, they compare apples (stocks) to oranges (bonds) like equals)

  2. dave commented on Apr 3

    Also, what are the quality of those earnings. How much in charges and option expenses are ignored?

  3. mdbllbr commented on Apr 3

    The Fed Model is not a forecasting model. It’s an arbitrage relationship between future path of the short term interest rate and future earnings. It helps you to link your macro forecasts (GDP, Inflation, short term interest rate) with future stock market performance.


    BR: I am not sure that is the most precise usage of the word “arbitrage”

  4. Macro Man commented on Apr 3

    The Fed model says more about the price of bonds than the price of stocks. However, it does explain why buybacks are so prevalent; earnings yield on equity is higher than the yield that companies can get on deposit, in Treasuries, or, for investment grade corporates, from buying back their own debt.

  5. V L commented on Apr 3

    Did the model consider these facts?

    “… Share prices have been propped up by debt during recent stages of this rally. Debt has funded corporate buybacks of shares, and those buybacks have made specific stocks look more attractive (fewer shares outstanding means higher earnings per share, since the earnings get spread over fewer shares) and supported the stock market in general by reducing the supply of shares and adding to earnings growth. Debt has funded dividends. Yes, I know dividends are supposed to come out of earnings, but recently many companies have discovered that it’s easier to simply borrow the money in order to pay it out to investors. In that way, too, debt has supported stock prices by making dividend growth look better…”

  6. Macro Man commented on Apr 3

    Probably not, because it’s not true. US corporates have large net cash positions and financial leverage is low.

  7. wally commented on Apr 3

    To extend your point about double-counting interest rates: a rate of ‘0’ would obviously be ideal. The model does not consider that there may be reason why that rate might lead to other dificulties… say, a bubble economy, for instance.

  8. Winston Munn commented on Apr 3

    In my view there are two items worth investigating further: first, “profits are rising faster than shares” – this is a little misleading as buybacks have reduced shares and profits are not rising but beginning to trend downward; second, “bond yields shows confidence that inflation will stay in check” – the ability of the bond market to predict inflation may well have been substantially reduced by the demand for bonds created by the negative trade deficit.

  9. Josh commented on Apr 3

    I find it interesting that 3 companies that manage $1.3 Trillion are going out of their way to tell Joe Retail that stocks are cheap and he should buy. What would these companies have to gain from this?

    And of course if you name drop, like Fed Model, you are sure to get some reaction. Forget negative savings, high inflation, decelerating earnigns, the myth of forward PE, and the tapped out consumer.

    The problem is the consumer is so tapped out, they are taking money out of the market (their savings plan called a 401k) and using it to pay their mortgage.

  10. V L commented on Apr 3

    “BlackRock Inc., Fisher Investments Inc. and Schroders Plc” are smart guys to really believe in the misleading B.S. they are spreading. (They know it)

    It makes me wonder if it is an exit strategy for them. (Dumping their shares to Joe)

  11. Michael Schumacher commented on Apr 3

    I guess we had to have a comedy piece did”nt we??

    Seriously what’s missing in any analysis is the quality of earnings, if a company beats expectations it seems no one looks to the real reason like say a tax benefit or similar. Just that they “beat”….I see no reason why that trend will not continue. Fisher, Draper et al have quite the motivation to be bullish since bad news does’nt sustain money flow into thier funds.

    AS someone posted previoulsy companies will make the numbers, especially in the short term since all they have to do later on is say we’ve uncovered “accounting issues” and then re-state…and the stock acts as if the “fakey” numbers were real even after they’ve stated the truth- well what they say is the truth.

    Pretty disturbing also how all of a sudden the market soars (futures-at least) on on “easing of tensions”…too bad it did’nt drop much on the escalation of the same “tensions”…….

    Spin in full effect..


  12. traderb commented on Apr 3

    “Europe has eclipsed the US in stock market value for the first time since the
    first world war in another sign of the slipping of the global dominance of
    American capital markets. Europe’s 24 stockmarkets, including Russia and
    emerging Europe, saw their capitalisation rise to $15,720bn (€11,819bn) at the
    end of last week, according to Thomson Financial data. That exceeded the
    $15,640bn market value of the US.”

    From Morgan Stanley today

  13. Peter commented on Apr 3

    I’m glad you touched on the fed model as I read the Fisher piece too. If they had used the fed model in 1981, stocks with depressed earnings and sky high interest rates, they would have declared how crazy, expensive stocks were just as the greatest bull market in US history began.

  14. Jay Weinstein commented on Apr 3

    By far the best work done on debunking the idiotic [and it is idiotic] Fed model has been by Cliff Asness at AQR. Go to their website or google him to find his research.

    My much simpler retort is always as follows: If long rates dropped to 1%, does that mean stocks should trade at 100 PE and equities would make tons of money? Of course not–just ask the Japanese.

  15. Barry Ritholtz commented on Apr 3

    BTW, I think the quality of earnings is better than they have been in many years . . .

  16. Adam commented on Apr 3

    Barry, I’ve said it once on these boards and I’ll say it again: if you wish to do your readers a great service, discuss current P/E ratios in terms of P/multi-year moving average of E. Graham suggested using no less than 7 years of earnings; Robert Shiller suggests 10 years. This method smooths out brief dips or spikes in corporate earnings and gives you a better sense of where valuations are.

  17. Alex Khenkin commented on Apr 3

    1. Bonds do not “predict” inflation, they follow it. Obviously.
    2. Comparing “earnings yield” (a number on a piece of paper that may or may not represent real cash, may or may not find its way to the shareholder’s pocket, and may or may not be there tomorrow) with a bond dividend (a guaranteed cash payout on a guaranteed principle investment) sounds exactly the way it sounds – rather ignorant.
    3. A technical point could also be made that using an investment with one duration as a benchmark of an investment with a considerably different duration is invalid.
    Small Investor Chronicles

  18. Michael Schumacher commented on Apr 3


    I can’t disagree with your statement that they have been better however one must look at where you started from to arrive at that opinion. If earnings quality is at, say 10%, (just using numbers to illustrate) and it “rises” to 20% the percentage increase is wonderful however 20% is still pretty poor overall and does’nt really support that they are any better….all in where you begin.

    How was the game?


  19. Philippe commented on Apr 3

    At this above address you may find a documented analysis on corporations borrowings as compared to capital expenditures or alternatively stocks purchases.
    Equities are not expensive if purchased through derivatives and require more thoughts when paid in cash.
    This article of Bloomberg was published around….April 1ST day where you may have more imagination than reality oblige.

  20. Fred commented on Apr 3

    Bashing this model is akin to bashing a PEG ratio, or PE analysis, etc. It cannot be viewed in a vacume. It is a relative value tool — showing where $$ is being treated the best. As Macro Man suggests, this explains why there is so many buybacks. It certainly is not a timing tool…but looking back on the bubble, it was obvious that bonds were a better deal than stocks at the peak! You couldn’t give away zero coupon bonds in 2000!!!

  21. glenn_in_MA commented on Apr 3

    John Hussman also does an excellent job at debunking the Fed Model.

    BTW, John relates current P/E ratios in terms of Price to Peak Earnings

  22. S commented on Apr 3

    As Alex implied, a much better analytical measure to compare equities with corpoates is free cash flow yield per share.

    On another note, I wonder how undervalued the Nikkei is using the FED model, with 10-year JGB’s yielding 1.7%?

  23. lloyd commented on Apr 3

    P/B ratios sure don’t look cheap and insiders are dumping stock at close to the highest level ever. Gee….I wonder why?

  24. Michael Schumacher commented on Apr 3

    Today’s “rally” should be named the David Lereah Market Assumption rally……

    No one reads any longer…..


  25. mark commented on Apr 3

    I dont know but this market is rockin n rollin today. Up there near the highs before that february 28 selloff. wow! How high can we go?????!!!!!! awesome!

  26. jack commented on Apr 3

    Isn’t april strong for stocks due to IRA contribution deadline?

  27. Michael Schumacher commented on Apr 3

    re: IRA contributions….

    Yes I’ve heard that too but I file that in the same category as “tax selling” in Decemeber. It means very little overall…..certainly not to power this “rally” or whatever we call it now.


  28. Fred commented on Apr 3

    Sentiment is about as bad as it gets right now. Do you really think housing and subprime slime is NOT baked into the sentiment cake? Is there really ANY news there? When the sentiment gets this bad (COT, short interest, put call, etc) it’s like stealing sweets from a child.

  29. Michael Schumacher commented on Apr 3

    Sentiment is not “as bad as it gets”…how else do you explain market movement, especially like today?????

    If it were truly “as bad as it gets” we would be quite lower and definatley NOT challenging an all time high in the Dow. This smacks of creating the space for the begining of earnings season- that just screams shortfall- as at the very worst the market was down “only 2.5%” at it’s worst level (just below 12k). And then you’ve got people, like yourself-Fred-, saying that all the warnings are already baked in the market. If we had a real correction, say 10% or so and we’ve rallied back half of that then I would agree that sentiment is fully baked into the market. But lets just continue to ignore the reality of the housing market, labor, inflation, MEW’s, and other assorted ways to allow sentiment to be “as bad as it gets”…..thanks I needed a good laugh this morning.


  30. Fred commented on Apr 3

    I’m amusing you? Interesting sentiment there.

    When I say sentiment is bad, it is a function of $$ BETS PLACED…not blog ranting. You subscribe to the majority (media fed) view that I’m happy to bet against.

    Good luck (again).

  31. Fred commented on Apr 3

    Further…remember that many nice moves in the equity market occur as earnings slow down…(baked in)

  32. Michael Schumacher commented on Apr 3


    Yes you are amusing in the sense that you don’t even try to argue. You just offer up another “clarification” of your “position”.

    and just so ya know…..I’ll repeat here what I posted to the other “fellow”. Confusing trading styles and opinions or “blog writings” is somehting you have excelled at. But I await your next “clarification”…..

    Try participating in an arguement instead of repeatedly “clarifiying” your original position. I’m sorry…..”it’s different this time” was your last “position”


  33. Fred commented on Apr 3

    There is nothing to argue about Michael. I am referring to published (read: factual) readings on sentiment. Are we clear? Let me say it as clearly as possible — the statistical sentiment is at an extreme (bearish). No opinion or spin in this statement, nor fishing for an “arguement”.


  34. Michael Schumacher commented on Apr 3

    I totally disagree with you however don’t you think you should have framed it that way in the first place??

    Betcha did’nt.


  35. Fred commented on Apr 3

    Is that a survey?

    Here’s another one (from

    Greenwich Alternative Investment survey reported 69% of macro hedge fund managers expected the S&P 500 to decline in April (the highest level of bearishness since the survey began in January ’04)?

    I don’t like crowds.

  36. flipper commented on Apr 3

    I suggest that anyone should read an article named “Buble Logic”, the author is Clifford Asness.

    The later and revised version can be downloaded from CFA institute website.

    The author examines how P/E ratios are set with data starting with the late 19th century.

    It is shown that “rate diferential”, namely E/P – bond yield is relevant to future returns only due to E/P, what is the fed model is flawled, and it’s only P/E and it’s inverse – “earnings yield” that has significance for future returns.

    Also it’s is show that different generations of investors set multiples according to their experience – past volatility of stock market plays an important role. The bottomline is that investor who survived the buble crashing will never set multiple to high again. It’s the new and unexpirienced generation which will eventually create a new buble

  37. Jdamon commented on Apr 3

    Flipper, I think you are spot on regarding the p/e multiples. It will be a long time before multiples approach the 30+ avg. they were in 2000 before the tech crash.

    In my opinion, this is what has created the “undervaluation” of stock prices that bulls have been pointing out (leading to record levels of M&A activity).

    Hard to argue that relative to any other asset class out there that stocks are in any way shape or form overvalued right now.

    Do you think a recession, which you only know you had after you had it, will bring valuations lower than they are now? I think the market has already priced in much slower earnings growth, as well as GDP growth. This is why we will eeck out a 7 – 9% increase in the overall market this year.

  38. Grodge commented on Apr 3

    One interpretation of such a valuation model would say that stocks will continue to follow this longer term trend of “under-valuation” until some event, such as a rapid capitulation, turns it around.

  39. flipper commented on Apr 3

    Jdamon, investors tend to make common mistakes, for example the multiples are set lower during inflationary period, while stocks are clearly an inflation hedge.

    I think that the only reason for problems now is potential credit cruch, but that is something most people have been worring about back in 1987, and nothing has happened yet, if it finally comes, shure stocks will be depressed.

    Lloyd, o good point on earnings – they can not grow at double digit pace for a long period of time, in long run they cannot grow higher than nominal gdp. Ofcourse we are talking world economy now, since half of snp500 earnings comes from abroad, but even so, continuing growth at same pace seems unlikely.

  40. flipper commented on Apr 4

    Barry, another thing on fed model and your post update.

    The fed model is clearly rubbish, since it compares yield on high risk residual claim with a riskless goverment bond.

    It’s an ABC of stock valuation, that a stock discount rate should include a risk premium.

    And about 1981 – if you look at multiples during depression or recession, it’s not the current earning you should be loking at but the peak earnings since in long run they shouldd recover to those levels and eventually exceed it.

    Also that period had high inflation and hence high bond yields, and that investors failed to recognize, is that stocks are inflation hedges, and that inflation (and high rates) is temporary.

  41. Macro Man commented on Apr 4

    Flipper, the reason that P/E’s tend to be low during high inflation periods is that the discount factor applied to future earnings streams is high, which makes stock prices appear ‘low’ relative tocurrent earnings.

    In a zero inflation world, futures earnings streams what have an identical present value to future value, and thus stock rices would appear ‘rich’ compared to current earnings.

    If you want an inflation hedge, buy commodities or TIPS.

  42. flipper commented on Apr 4

    Macro Man, i am aware of that, but historicaly , US companies were able to pass 94% of inflation to consumers.

    So, while imperfect, they are really an inflation hedge, while not so good as commodities.

    And while the discount rises, so do nominal earnings.

    Shure, every company has it’s own unique ability to pass inflation to consumer, but studies show that is overall relatively high > 90%.

    take the simpliest DCF, that is P/E = g/(r-g).

    If we ignore small order terms and add an increase in inflation, that becomes

    P/E = (g + l*i )/(r+ i – g – i*l), there i is an increase in inflation and l is inflaton flow-throw to consumer. If you plug there historical estimate of 0.94, you get a lower multiple, but not dramaticly lower as it has been in late 70ies, ofcourse it all depends on exact figures.

    Also the impact is clearly different for different sectors – that is financials which have a lot of credits with fixed rate will be hurt the most, but many of their debtors with high ability to pass inflation are in beneficial position.

    The very low multiples and “equities are dead” articles are just evidence, that a crowd almost always gets it all wrong, like Nasdaq buble is.

    But you should be able to stay solvent during those times :)

  43. Barry Ritholtz commented on Apr 4

    As I specifically quote above, according to the Fed Model, in 1981 stocks with depressed earnings and sky high interest rates were crazy expensive — just as the greatest bull market in US history began.

    That is a VERY significant flaw in the Fed Model. Missing what may very well be the greatest buying opportunity in a generation raises some very serious questions about any model that proclaims to be able to determine valuation.

    In the pre-1982 version, interest rates were inordinately high and profits inordinately low, thus stocks looked extremely expensive. A bit of mean reversion on both of those, and stock valuations suddenly became much more reasonable.

    Today, we have the exact opposite — the 10-year interest rate level is still historically very low, while profits are very high. A little mean reversion on each and sudddenly stocks will look pricey.

  44. Macro Man commented on Apr 4

    In fairness though, Barry, the total of return of bonds was broadly similar to that of equities from the end of 1981 through the end of 1986. Stocks obviously raced ahead in 1987 but ended the year with only a moderate aggregate outperformance over bonds. And naturally, on a risk adjusted basis, bonds were a far superior investment to equities until the late 1980’s.

    Equities and bonds were both cheap in 1982, and both delivered superb nominal returns for the ensuing decade. The “Fed model” did correctly spot the outperformance of bonds on a risk-adjusted basis, however.

    Today, one could make a credible argument that neither stocks nor bonds are cheap…but of the two, equities appear to be the cheaper, on the basis of that model.

  45. Barry Ritholtz commented on Apr 4

    I do not ever recall hearing that the Fed model nows forecasts risk adjusted returns — that sounds like a very much after the fact revision.

    The reality is that the Fed Model would have missed the greatest buying opportunity in our life time. That alone causes met to question its utility . . .

  46. Macro Man commented on Apr 4

    OK, but by the same token it got you in the greatest bond rally in history, which made you just about as much as the equity rally for 6 years. Where is the claim made that the ‘Fed model’ attempts to model anything but relative value? Certainly not by Blackrock, Fisher, or Schroders above!

  47. Barry Ritholtz commented on Apr 4

    Actually, they are saying BUY STOCKS — not BUY STOCKS AND BONDS.

    They are using the Fed model to persuade the purchase of equities.

    I am saying that it is not the best measure for when to buy stocks — and may even be a very poor measure!

  48. Macro Man commented on Apr 4

    They are saying equities are cheap RELATIVE TO BONDS. Insofar as people want to put their money somewhere other than cash, that’s not an unreasonable proposition.

    Insofar as people wanted to put their money in something other than cash in 1982, bonds were not an unreasonable proposition, either.

  49. Fred commented on Apr 4


    Looking at Hays Advisory’s version of this model, STOCKS WERE A BETTER BUY in 1981. In fact, the only time bonds have been a screaming buy (relative to bonds) was 1999 – 2000. Not a bad call, no?

    Their definition:

    “This measures market valuation by analyzing where money is being treated the best by the financial markets. The S&P500 Forward Earnings Yield is the consensus earnings estimates expected for the next 12 months divided by the S&P500 price index. Then a simple comparison can be made to determine whether stocks or bonds are more attractive at any particular point in time.”

    Additionally, the IBES Valuation model has stocks 31.4% undervalued (currently) to bonds. It’s NOT saying stocks will go uo 31%… bonds yields can obviously go up to help aleviate the valuation difference.

  50. flipper commented on Apr 4

    Fred, on what data set they test the model?

    Fed model works faily good after 1960, but does a very poor job with data before that time.

  51. Fred commented on Apr 4


    I believe they use IBES estimates data, if that’s what you mean. Their data starts in 1979.

    See the definition above.

  52. Fred commented on Apr 4

    Ticksersense just posted on the weaker earnings growth, and the break of the record double digit growth streak. This is not something the bears should count as a plus:

    “While the 18 consecutive quarters of double digit growth is the longest streak of its kind, there was a 13 consecutive quarter streak from 1992 to 1995. Once that streak ended, growth remained low for the next couple of years, but it did not seem to negatively affect the markets.”

  53. Gordon Haave commented on Apr 4

    >The reality is that the Fed Model would have >missed the greatest buying opportunity in our >life time.
    Barry, the way your comments are written, you are implying that the Fed Model would have kept you out of stocks for the 80’s. The Fed Model did not say you should stay out of stocks for the entire boom of the 80’s. At most you would have missed a year or so – and made great money in bonds in the meantime.

    There are a multitude of other comments on this thread that are also unreasonable. A common theme is: Well, you are comparing risky stocks to riskless bonds, they should have a higher earnings yield. OK. All that says is that the key determinate for relative cheapness/expensiveness is not when one is greater than the other, but rather when there is a risk premium built in for stocks.

    Further, the argument is made that, well, present earnings don’t properly predict future earnings. Well, OK then, adjust predicted earnings on your own.

    All that is meant to say that there are very few black and whites in investing, and many models might tell us quite a bit even if they are imperfect.

  54. Barry Ritholtz commented on Apr 4

    I’m not sure about the entire 80s — just the best entry in 1981.

    But I’m not sure when the Fed model would have gotten you in post-1982, considering its a relative valuation model.

    I’ll do some digging/asking around . . .

  55. M CAM commented on Apr 8

    The US equity markets are awash in cheap cash! The only thing that can send this market lower is repeated fed hikes. Too much cheap green chasing to few opportunity’s. Turn off the printing press!!!!!

  56. A Dash of Insight commented on May 7

    Critics of the Fed Model

    Yesterday’ we took note of the Bloomberg story, Cheapest Stocks in Two Decades Signal Bull Market, which quoted several large and successful fund managers. This story stimulated a post from Barry Ritholtz at the Big Picture, where he said that

Posted Under