Challenge, Prove or Disprove

Historical Equity Premium of Stocks over Bonds

I hypothesize that at least 50% of the historical real total return premium of stocks over bonds since 1933 is due to the actions of the Fed in driving monetary supply inflation and alternately supressing and boosting interest rates. It is certain that stocks and bonds will not be equally sensitive to Fed actions over both the long and short term. I suspect that some potential reader will be especially qualified to prove or disprove this hypothesis through simulation.

The basis for this hypothesis is that once purchased, a bond has no means for combatting price inflation, while stocks can partially capture an inflationary trend by buying production factors earlier and cheaper and selling finished products later and dearer. Also, stocks will be artificially inflated in value at times when the Fed is supressing interest rates, and the discount rate for valuing future free cash flows is low.

There are any number of possible simulations that might address this issue, but a quick and dirty skeleton of an example is as follows :

Start with equal dollar amounts of the S&P500 and 20 year treasuries.

Set up three 20 year trajectories for Fed action, resulting in 2.5% price deflation, no price inflation, and 2.5% price inflation, all annual results. Both price inflation/deflation and short term interest rates need to be assumed, hopefully in a somewhat consistent manner.

Compare both the stock and bond portfolios at the end of 20 years for each of the three Fed assumptions.

The total return of the bond portfolio for 20 years will be defined as the real value of the final account value, including all of the coupon payments reinvested in 90 day treasuries at the existing interest rates encountered along the way, and adjusted for overall 20 year price inflation/deflation.

The total return of the stock portfolio will be the real account value at the end of 20 years, liquidated into cash and adjusted for price inflation/deflation. Dividends along the way will be invested in 90 day treasuries just like the bond coupons above. Assume nominal dollar earnings and moderate growth, and adjust for price inflation over each 1 year production cycle.

To smooth the result against interest rate variations at the end of the 20 year period, liquidation will start early.

After 16 years, 20% of the index fund value will be converted into the equivalent of 4 year treasuries.

After 17 years, 25% of the remaining index fund value will be converted into the equivalent of 3 year treasuries.

After 18 years, 33.3% of the remaining index value will be converted into the equivalent of 2 year treasuries.

After 19 years, 50% of the remaining index fund value will be converted into 1 year treasuries.

After 20 years, the entire remaining index fund will be liquidated to cash.

This is all just a suggestion. Anyone that might be interested is encouraged to report their simulation setup and results.



Share this

While it is certain that

While it is certain that stocks should prove to have some natural hedge against inflation (as prices rise so should be P in the P/E ratio and on the assumption that companies can pass though price increases to consumers, hence hedging themselves against cost inflation)

But you miss the key reasons for the difference between the returns you site. The primary one being the difference in risk between the two investment choices. Government bonds are inherently less risky and therefore demand less of a risk premium (i.e. return) vis stocks. A second reason is the bias towards the successful: companies that perform well grow in terms of composition of the index in question and those that do not, and eventually drop out of the indices (delisted, bankrupt) altogether. Only sudden reversals of fortune (as we saw with the nasdaq from 2000) make a significant dent in index performance. The S&P 500 contains a select group of performers. Losers drop out, newcomers on a growth trajectory get into the club and bias returns higher. I would point out that a new member to the S&P almost always gets a fillip from funds forced to buy (index funds, any that are benchmarked against the index) and this also skews restults. As a former portfolio manager myself I have seen just this kind of action in action.

You would do better to compare corporate bonds with stocks since the risk set are more equal (though of course bondholders are senior) but then you must make a volatility adjustment to the returns: the resultant sharpe ratio is the number you are looking for: the higher that ratio be more efficient the portfolio in question and the better overall performance (in risk adjusted space) one is over the other.

All, Reference


Reference added.

Regards, Don

What kind of software would

What kind of software would one use to do that?

Jacob, What kind of software


What kind of software would one use to do that?

I don't know. Maybe APL or a spreadsheet. Extra credit if you use IBM1620 machine code with a punched card deck.

Regards, Don

I agree that the comparison

I agree that the comparison is meaningless with adjusting for risk. However, I've always heard the paradox as "stocks perform better than bonds on a risk-adjusted basis", so it will probably still hold.

The explanation that stocks can better deal with inflation doesn't solve the paradox, it just begs the question of why investors don't demand higher rates of interest on bonds because of the inflation risk.

Patri, It's simple really:


It's simple really: Because the expected return on the vast majority of bonds is pretty well established a priori while for stocks it is entirely uncertain. In US government bonds the investor knows with great certainty what the total return to maturity will be.... yes, of course selling said bonds before maturity may entail a loss (based on duration and changes in prevailing interest rates) but holding to maturity provides a given return. For this certainty, the investor is happy to part with some other possible, theoretical, better return.

In other words, the downside risk in most bonds has a floor, providing no default (which is something that US govt bonds seem to provide an effective hedge agaist). Corporate bonds come closer to including some risk of non-payment, but such events are really rather rare and equity holders remain second-class creditors. The seniority structure alone should count for some of the difference.

And on average, the investment community on aggregate has to be more concerned with downside risk than upside opportunity. (This also contributes to your skew and the seemingly incomprehensible spread. Hmmm... I may have to blog something on this from my own experience....)

Now if you go back an do your 20 years of stocks against, say, emerging markets bonds, you find something rather remarkable. While the latter does entail a certain default risk, total returns for the sector have, over-all, been better than stocks and on a sharpe ratio basis (adjusted for the volatility of returns) beaten the pants off stocks. But no one would suggest everyone dive in to that hairy sector.....

most finacial sales

most finacial sales professionals for various reasons are forced to use gov't statistics which are criminally/purposefully biased towards low inflation, even though anecdotal information shows us that they are practically worthless as investment advice .outside the box thinkers know this. almost all of the much vaunted stock gains can be accounted for by inflation and survivor bias. the hedonic deflation lies perpetrated by the Greenspan posse are the most laughable and sad. i'm waiting for the day they modify or eliminate the crb index to more 'accurately reflect real world conditions'. that'll be the 'bell ringing' you'll need to get everything you've got into gold bulion.