Rethinking Stocks for the Long Haul

[samclem]

From the article:

Quote:

especially with the market up 90 percent since reaching its low on Mar. 9, 2010.

I don't know what "market" he's referring to, but if it is US equities I think he meant 2009. Maybe Business Week gave their editors the week off.

Wilshire 5000 Total Market Index (full-cap)

Anyway, yes stocks might be risky (i.e. volatile, which, IMO does a poor job of capturing the "risk" most people need to be concerned about). What alternatives does the author propose--lock in those appetizing 2.5% CD yields for 5 years and wait for the coming 7% inflation? Maybe some rock solid California bonds? It's very possible that the equity premium in US staocks will be lower for the next 50 years than the last 50--that's a good reason to diversify globally and to own US multinationals. Is it likely that over a multi-decade period stocks (with higher risk) will yield less than corporate bonds? Most importantly: will a portfolio with equities, bonds, and cash that is rebalanced periodically perform better or worse than a portfolio without equities? Which approach is truly riskier (not just less volatile, but also less likely to lose to inflation?)

Quote:

Originally Posted by harley Just another paycheck article, IMO.

+1. But, hey, what is he supposed to do? The deadline is approaching, the boss wants 1000 words to fill the page . . .

[GregLee]

Quote:

Originally Posted by traineeinvestor I have mixed views about the article but on the whole found it unconvincing.

I found it completely unconvincing. Consider this passage:

Quote:

Like their scholarly peers who are more positive on equities, these two believe bull markets and bear markets largely cancel each other out. In other words, stocks are "mean reverting" over the long haul. It helps explain why U.S. stock returns have been so consistent over time, about 7 percent after inflation since the early 1800s.

A real 7 percent gain over more than 200 years demonstrates stocks are "mean reverting", or bull markets are largely canceled out by bear markets? That's not even close to making sense.

[FIRE'd@51]

Quote:

Originally Posted by FIRE'd@51 Maybe not. I believe the answer may lie here

OK, I didn't really expect people to run off and read the Margrabe article thoroughly (I linked it as a kind of swipe at MB who seemed to suggest we couldn't be serious ). I think it is relevant, however, and here is my interpretation of how it applies. I think the most fascinating point in the article linked by MB is that raised by Zvi Bodie near the end when he talks about "shortfall insurance".


— If stock market risk fell the longer the investment was held, the cost of owning options to protect the portfolio from a "shortfall loss" should also decline. (A shortfall loss is a loss relative to what an investor could have earned by putting money into a safe asset over the same time period, such as U.S. Treasuries.)

In an elegant exercise in 1995, Professor Zvi Bodie of Boston University tapped into the insights of option-pricing theory to challenge the notion that stocks became less risky for the buy-and-hold investor. Yet the cost of insuring a portfolio with options goes up, not down, with time. For instance, he estimated at the time that a one-year shortfall "put" option would cost 7.98 percent of the portfolio, and at 30 years the cost of the shortfall insurance jumped to 41.63 percent. (At 200 years the comparable number was 84.27 percent.)


Bodie's numbers can be computed using Black-Scholes (which is really a special case of the Margrabe option, i.e. an option to exchange a zero coupon Treasury security (i.e. the strike price) for the S&P 500). A one-year call option on the total return of the S&P 500 with a strike price 5% (the risk-free rate at the end of 1995) above the then value of the S&P 500 would cost 7.98%. So if the S&P were 100 at the end of 1995, one could exchange the strike price 105 for the index itself plus the dividends paid in one year. Bodie talks in terms of a put, which is the cost of the shortfall insurance, but since the strike price is the forward price of the total return of the S&P 500, the call and put prices are the same. Bodie then says that if you price the cost of the shortfall option for 30 years, it is 41.63% of the initial value of the S&P 500. That's because the strike price in the 30-year example would be 432.19 (the future value of a risk free asset that grows at 5% per year for 30 years.

Bodie then argues that, because the insurance cost (i.e. the put price) against a shortfall increases with one's time horizon, the risk of underperforming the risk-free asset must be increasing as well. I think this is a very clever and interesting insight, since common "wisdom" is that the risk of holding equities decreases if one increases one's investment horizon. Bodie seems to be saying this isn't really the case, since the risk of a shortfall increases with horizon, and it's shortfall risk one should focus on.

[HFWR]

Quote:

Originally Posted by GregLee A real 7 percent gain over more than 200 years demonstrates stocks are "mean reverting", or bull markets are largely canceled out by bear markets? That's not even close to making sense.

I assume the author considers seven percent to be the mean to which bull and bear markets revert, thus canceling each other out...

That stocks are risky should be crystal clear at this point. I was diversified across countries, and asset classes, and when the big dump came, all correlations mean-reverted to one...

[samclem]

Quote:

Originally Posted by FIRE'd@51 Bodie then argues that, because the insurance cost (i.e. the put price) against a shortfall increases with one's time horizon, the risk of underperforming the risk-free asset must be increasing as well. I think this is a very clever and interesting insight, since common "wisdom" is that the risk of holding equities decreases if one increases one's investment horizon. Bodie seems to be saying this isn't really the case, since the risk of a shortfall increases with horizon, and it's shortfall risk one should focus on.

Can you help my simple mind with this:
1)The article's premise is that the market is fundamentally overvaluing equities (relative to bonds).
2) Brodie's research indicates that the market "believes" (as expressed by long-term put prices) that the risk of equity underperfomance (vs risk-free investments) increases with time.

If "the market" can be wrong about the correct price for equities today (#1 above) is it proper to base another leg of the argument (#2) on what the market believes about the future price of equities vs. a risk-free investment? Seems to me we'd need more data to determine in which case the market is making a more accurate pricing determination.

[Nords]

Quote:

Originally Posted by FIRE'd@51 Bodie's numbers can be computed using Black-Scholes (which is really a special case of the Margrabe option, i.e. an option to exchange a zero coupon Treasury security (i.e. the strike price) for the S&P 500). A one-year call option ...

Aw, shucks, I was with you until you started using Black-Scholes as an example of an appropriately-priced option. Now you're gonna spawn a whole bunch of comments about the market being correctly priced or efficient or some propaganda like that.

We haven't even gotten into the wisdom of dividend stocks, let alone commodities. Why, you could pay $7000 today for a new koa planting on the Big Island and pull in a cool $280K in just three decades.

[clifp]

Quote:

Originally Posted by traineeinvestor I have mixed views about the article but on the whole found it unconvincing. I am much more comfortable buying equities for the longer term than I am buying bonds - given (i) that the earnings yield (inverse of PE) is higher than the yield on longer term bonds (YMMV - depending on which equities and which bonds you are talking about) and (ii) I belive that corporate earnings as a whole should grow over time (in both nominal and real terms), in my rather simplistic view of the investing universe, equities are actually safer preservers of wealth in the longer term than bonds. Given that I expect we will continue to see at least modest inflation in the forseeable future, this would hold true even if corporate earnings only grew in nominal (not real) terms. I appreciate that in the shorter term (or longer) there will be occasions in which equities do very badly - but that does not destroy the longer term case for owning equities. Instead, it leads to the important conclusion that one shouldn't hold all equities...

I think you are well past the trainee stage as an investor I very much agree with this post.

If the articles point is that even over long 20+ years periods of time stock maybe a poorly performing asset class, and if you have bad timing for retirement you can be in trouble. Then my response is well .

If the article's point is that stocks in the future are going to underperform bonds (especially treasury bond) then I just don't buy it and the author certainly didn't prove it.


I think there are fundamental reason that equities will outperform bonds and it has to do with perceived risk plus a fundamental ignorance among most people about investments, coupled with my believed that continued productivity gains will increase corporate profits.

There are two ways that any assets will generate returns for investors: the greater fool theory, and an fundamental increase in the future earnings potential of an investment. The vast vast majority of financial reporting, and discussion focus on the greater fool theory, p/e expansion, earnings momentum, investor confidence, future interest rates etc., the 20 something who struck it rich by flipping house, bond trading billionaires etc. Fundamental improvements in an asset, e.g. the new Pampers which are more absorbent, and requiring 10% less material to make and are built in a new factory that requires 1/2 the labor, are generally ignored.

Over the last decade just as back in 99 and 2000 many people
believed that stocks only went up, but for every person who was guilty of this fallacy (including myself ) there were several people who felt the same way about houses. The reality reset of 2008 and the overall lost decade of stocks shattered most people illusions about equities, and probably about houses. It is entirely possible that the average person is going to be scared off from investing in equities for a long period, perhaps in many cases for the rest of their life, like what happened to many of the "greatest generation" who swore off stocks after the depression.

This very well may cause a decrease in demand for equities that might persist for many decades and perhaps would cause returns to drop below the historical 7% or so average. However, stocks don't exist in a vacuum many people were similarly burned in real estate. Imagine a person who saw his 401K drop from $401,000 to $201,000 and then recover to $300,000, at the same time he saw his $500K house that he put $100K down, drop to $300K and now $250K and is walking away from it. Which is going to be the more traumatizing? the $100K loss of his down payment and the ruining of his credit score or the $100K paper loss in his 401K balance. I am not sure which is going make the investor more gun shy, but I think that a future real estate investment would be scarier.

Right now most of the scared money is in bonds, 0% interest money markets, and 1-2% interest rate CDs. I believe a few more European countries defaults along some big cities in the USA, plus some losses for California and Illinois bond holders will be sufficient to traumatize bond holders.

It is entirely possible, I'd even argue likely, that many investors by 2012 will have managed to hit the trificta losing money in stocks, real estate, and bonds. Given the alternatives stocks stop looking that scary on a relative basis.

However, all of this discussion is once again focused on the greater fool theory, what sucker can I sell my asset to so I can make money.

The Buffett/Graham quote "In the short run the stock market is a voting machine in the long run it is weighing machine.", applies to all capital market not just the stock market.

On a fundamental level, corporations/stocks look significantly better than they did a decade ago and relative basis much better than bonds, real estate or commodities. I don't think anyone would argue that ability of bond issurers (the vast majority being government not corporate) are in better position to repay their debt now than they were a decade ago. I am no real estate expert, but I suspect the future prospect of real estate over the next 10-30 years looked better in 2000 than they do today.

In contrast the fundamentals of corporations look very good. They have strong balance sheets, many have record profits, and productivity gains have accelerate over the last decade.

Most importantly while they have probably lost 30 million (un and underemployed) consumers in the US and equal number in Europe in the last decade, they have picked hundreds of millions of new consumers (maybe even a billion) in Brazil, Russia, India, China. These BRIC consumer while they can't individually buy as much stuff as laid off auto worker or outsourced IT guy, collectively more than make up for their lost purchasing power.

Quote:

Then again, I'm something of an equity cultist.....

Me too because the only way for me to keep up with the emerging markets is to have ownership of companies who provide these new consumers with goods and services.

[GusLevy]

Some partially random thoughts near midnight:

1. In my opinion, the article linked in the original thread makes the simple argument that most investors underestimate the variance of returns of the broad stock market (S&P500) which then leads them to underestimate the actual length of time that often is needed for the S&P500 to "make-up" for particularly bad performance periods. I believe that the authors are concluding that the "mean-reverting" S&P500 long term return of 7% - which may or may not be true - oftentimes may take a lot longer than 10, 20 or even 30-yr time periods to ultimately revert (if in fact it really does at all). Therefore, investors who simply believe that any 10, 20 or 30-yr holding period for the S&P500 will guarantee themselves a reasonable approximation of the "mean-reverting" return of 7% or 8% or whatever return exists in their mind's eye are overestimating the actual odds of that outcome.

2. "In an elegant exercise in 1995, Professor Zvi Bodie of Boston University tapped into the insights of option-pricing theory to challenge the notion that stocks became less risky for the buy-and-hold investor. Yet the cost of insuring a portfolio with options goes up, not down, with time. For instance, he estimated at the time that a one-year shortfall "put" option would cost 7.98 percent of the portfolio, and at 30 years the cost of the shortfall insurance jumped to 41.63 percent. (At 200 years the comparable number was 84.27 percent.)"

I am only vaguely familiar with Zvi Bodie but when I read this paragraph in the article something didn't feel right. While I do myself generally agree with the authors' and Bodie's conclusions that most investors do in fact overestimate the actual long-term return and underestimate the actual volatility risk of the S&P 500, my hunch is that Bodie - while correct in his hypothesis - could not possibly be correct in his methodology since he implies using the zero coupon curve in his model (via Black-Scholes) but is then able to calculate his "shortfall loss" curve out to 200-yrs. The zero-coupon curve only goes out to 30-yrs so either he has a fundamental error in his analysis - no matter what discount curve he uses it wouldn't give him data on 200-yrs - or his analysis is not only really, really, really elegant but perhaps also magical.

3. a) Synthetic Long Stock = Long Call - Short Put
b) At-the-Money Long Call - At-the-Money Short Put - Stock = 0

I would never have the patience to do Finance as a higher academic endeavor but whenever I think about the synthetic stock equation along with the put-call parity it feels somewhat intuitive that stocks should indeed go higher most of the time, ceteris paribus, since options are priced on yield curves or discount rates that are positively sloped most of the time due to the simple concepts of time-value-of-money and general monetary inflation. So, as long as growth exceeds the discount or borrowing rates, stocks should perform better than treasuries. And, really, growth should exceed discount rates or whatever borrowing rates are since the people borrowing at these rates have to grow the money enough to pay back their loans and make some money for themselves. Anyway, this is the status quo until exogenous or endogenous factors within a political or financial system make it difficult for borrowers to make money grow faster than at the rate one borrows it at. My intuition tells me that investors often underestimate the severity and length of some of these exogenous or endogenous factors which leads them to correctly believe that the status quo exists most of the time but, unfortunately, to incorrectly believe that it does so for longer periods of time than it actually does. How's that for a midnight stream of foolishness...

[chinaco]

There has been a lot of research and debate from the financial academics on this topic.

I agree with the basic message. One thing that has changed part of the "Stock Equation" is low/no dividends... they ain't what they used to be! Now we hope to get growth in the price of the equity... which is nothing more than somebody else bidding it the price up... he opposite of this is people dumping stocks in panic. Companies pay out little to nothing to owners.


The problem with bond rates (compared to stocks for investor) is that historically they have been kept relatively low (in the US) except for certain periods.


I do not believe many will argue with the basic phenomenon described... and many use these methods to try to counter the negative effects of it

- strategic allocation and rebalancing... to capture the gains.

- holding more bonds as one begins to consume their savings.


Who is going to disagree with this quote?

Quote:

History's message is that investors need to assign a much higher probability to the risk of an equity bust around the time they might retire. After all, the stock market has declined 20 percent or more every decade or so, on average. "You can wake up in the morning with half your wealth gone," says Kaplan. "The time horizon reassurance disappears."

[FIRE'd@51]

Quote:

Originally Posted by GusLevy I am only vaguely familiar with Zvi Bodie but when I read this paragraph in the article something didn't feel right. While I do myself generally agree with the authors' and Bodie's conclusions that most investors do in fact overestimate the actual long-term return and underestimate the actual volatility risk of the S&P 500, my hunch is that Bodie - while correct in his hypothesis - could not possibly be correct in his methodology since he implies using the zero coupon curve in his model (via Black-Scholes) but is then able to calculate his "shortfall loss" curve out to 200-yrs. The zero-coupon curve only goes out to 30-yrs so either he has a fundamental error in his analysis - no matter what discount curve he uses it wouldn't give him data on 200-yrs - or his analysis is not only really, really, really elegant but perhaps also magical.

As I interpret it, Bodie is saying "Let's hypothesize a world where the yield curve is a flat 5%, and stock market returns are normally distributed with a constant volatiliy of 20%". Is this the real world? No, of course not. But it's a framework in which one can use Black-Scholes to get a handle on what one means by long-term risk. Conventional "wisdom" would argue that, even in the Bodie world, the risk of owning equities instead of the risk-free asset goes down with an increasing investment horizon. Bodie is arguing that this is not the case, if you measure risk in terms of the cost of shortfall insurance (i.e. a put struck at the maturity value of the zero-coupon risk-free asset). If conventional "wisdom" is incorrect in the Bodie world, it must also be incorrect in the more complicated real world of non-flat yield curves and non-stable parameters of the stock market's return distribution.

[FIRE'd@51]

Quote:

Originally Posted by samclem Can you help my simple mind with this: 1)The article's premise is that the market is fundamentally overvaluing equities (relative to bonds). 2) Brodie's research indicates that the market "believes" (as expressed by long-term put prices) that the risk of equity underperfomance (vs risk-free investments) increases with time. If "the market" can be wrong about the correct price for equities today (#1 above) is it proper to base another leg of the argument (#2) on what the market believes about the future price of equities vs. a risk-free investment? Seems to me we'd need more data to determine in which case the market is making a more accurate pricing determination.

Bodie isn't doing this. See my post (#30) above.

[eridanus]

Black-Scholes only applies to delta and even then it fails in reality. (Try and find an equally priced call and put. It doesn't exist.) There are other factors at work in the real world, such as gamma and vega. Most importantly, for this article, is that the "risk-free" interest changes over time and is, of course, not truly risk-free.

[FIRE'd@51]

Quote:

Originally Posted by eridanus Try and find an equally priced call and put. It doesn't exist.

Sure it does, if they are struck at the forward price.

[samclem]

Quote:

Originally Posted by FIRE'd@51 Conventional "wisdom" would argue that, even in the Bodie world, the risk of owning equities instead of the risk-free asset goes down with an increasing investment horizon. Bodie is arguing that this is not the case, if you measure risk in terms of the cost of shortfall insurance (i.e. a put struck at the maturity value of the zero-coupon risk-free asset).

I'll take a look at his work so I can better understand it. It seems to me (based on your synopsis) he's doing what I describe--using the curve of future put prices to show that equity risk, as perceived by the universe of those buying and selling options, does not decrease with longer holding periods.

If that universe of people is not in agreement with the current assessment of equity vs bond risks as indicated by current prices, we're left to determine which "market" is wrong.

[ziggy29]

Quote:

Originally Posted by samclem I'll take a look at his work so I can better understand it. It seems to me (based on your synopsis) he's doing what I describe--using the curve of future put prices to show that equity risk, as percieved by the universe of those buying and selling puts, does not decrease with longer holding periods.

Well, if you use the idea that all empires crumble eventually and that fiat money of such an empire becomes worthless, then yeah, the expected value of a portfolio with a nearly infinite time horizon would approach zero, would it not?

[samclem]

Quote:

Originally Posted by ziggy29 Well, if you use the idea that all empires crumble eventually and that fiat money of such an empire becomes worthless, then yeah, the expected value of a portfolio with a nearly infinite time horizon would approach zero, would it not?

Yes. But Bodie's work reportedly compares the expected volatility of equities to that of "riskless" assets, and these riskless assets would also go to zero in a collapse. Maybe the problem is that long-term traders aren't accurately accounting for the "thin tail" potential for failure of these riskless assets as well, so the relative risk of equities is overstated.

It's that 20% scenario that Bernstein tells us can't be avoided through diversification. Start digging that bunker!

Quote:

History’s best-case scenario was the Roman Empire, which survived more or less intact for about seven centuries (if you ignore the odd sackings of the capital after 200 A.D.). A wildly optimistic historian might give us another few centuries of economic, political, and military continuity. Back-of-the-envelope, that’s about an 80% survival rate over the next 40 years. Thus, any estimate of long-term financial success greater than about 80% is meaningless.

[ERD50]

Quote:

Originally Posted by eridanus (Try and find an equally priced call and put. It doesn't exist.)

Oh, I mean : shhhhhhh! I'm making millions of dollars an hour arbitraging these, don't let the 'secret' get out!

Seriously - they are equally priced, but you do need to adjust for any in-out of the money shift, and the log-normal shape of the curve as the stock can't go below zero.

-ERD50

[Nords]

That's a great post, Clif. Thanks.

[eridanus]

Quote:

Originally Posted by FIRE'd@51 Sure it does, if they are struck at the forward price.

Explain this using what derivative, please. A forward contract? A forward start? A future? A forward option?

[calmloki]

Quote:

Originally Posted by FIRE'd@51 Maybe not. I believe the answer may lie here

You are a bad bad person. That is not the right way to start a morning.

You are all way too sophisticated for me - absent dividends ALL stocks look like they are just Pokemon trading cards. Only given value because other people want them. My simple mind wants to see why something is of value - can you eat it? Is it good? Does it make heat or cold? Does it keep the rain off?