Bernstein: A Decade of Super-Low Returns [1.4% for 60/40]

[berntson]

I just finished the first chapter of Bill Bernstein's new book. What a great read! Everyone should buy the book and read it. Now. Immediately. Don't finish reading this post. Buy the book and then come back. (No, Bill didn't pay me to say that. )

The first chapter ends with ten-year expected returns estimates for a number of asset classes. And they are not rosy. Take a 60/40 three-fund portfolio for example. Let's assume that the stock allocation is split 70/30 between domestic and international (this is roughly what Vanguard's target date funds do). The three funds are TSM, VXUS, and BND. Given Bill's estimates, a 60/40 three-fund portfolio can be expected to have a real return of around 2.5% over the next ten years [Edit: Sorry, but there was a bug in my initial calculation of this number. A 60/40 portfolio has an expected real return of about 1.4% given Bill's estimates. So his outlook is even more bleak. Thanks goes to asset_chaos for pointing this out.].

What do you think of Bill's expected returns? How do you plan to deal with low expected returns?

[InvestorNewb]

I don't see how anyone can reliably forecast returns.

Wasn't he recommending to stay out of REITs last year? They are up >16% YTD.

[schuyler74]

What are the reasons why he predicts a 2.5% real return over the next decade? Is it due to a fundamental shift in the world economy or something else? Because if it's based on things "being different now" than they have been, then why limit it to just the next decade and not the next 50 years?

[berntson]

I don't see how anyone can reliably forecast returns.

Wasn't he recommending to stay out of REITs last year? They are up >16% YTD.

Keep in mind that these are ten-year predictions. Anything can happen over the course of a few months or even a few years.

What are the reasons why he predicts a 2.5% real return over the next decade? Is it due to a fundamental shift in the world economy or something else? Because if it's based on things "being different now" than they have been, then why limit it to just the next decade and not the next 50 years?

Bill is using dividends, the dividend growth rate, and valuations to calculate expected returns. He is assuming that valuations will eventually mean-revert. Jack Bogle uses the same method when calculating expected returns.

[MP1233]

What do you think of Bill's expected returns? How do you plan to deal with low expected returns?

Seems reasonable given current conditions, but I think it also reasonable that the "unexpected" will pop up at some point in time and change the expectations. If low returns are realized, people that need more will need to take more risk, save more, or work longer. What else?

[matjen]

I think the big thing to remember with any of these predictions is that there is a range of outcomes. They are very useful for showing the area that is kind of likely but there is a decent chance we could do much better and a decent chance much worse.

[kenyan]

I respect his predictions (guesses) only slightly more than a few others'. He is going to be wrong, just like everybody else. Will he be less wrong? I don't really know. I don't do anything different - just follow my IPS. Granted, my IPS will undergo ten reviews/possible revisions over this timeframe, but I don't anticipate any major changes. We'll see, though, as I have to admit that's a pretty long time.

[steve_14]

Bill Be(a)rnstein predicting low stock returns? You don't say . I actually agree with him here - low interest rates = low stock returns. I'd flip the small and large cap values though:

All foreign 5%
US Large 4%
US Small 2%

I expect growth and value to have exactly the same risk/return profiles in real fund results. I don't care about sector funds.

[schuyler74]

What are the reasons why he predicts a 2.5% real return over the next decade? Is it due to a fundamental shift in the world economy or something else? Because if it's based on things "being different now" than they have been, then why limit it to just the next decade and not the next 50 years?

He's using dividends, the dividend growth rate, and the assumption of mean reversion to historic valuations. Jack Bogle uses this method in print and in interviews when calculating expected returns.

Right, that makes sense. As I understand it, the general equation is:

RealReturns = Dividends + DividendGrowth - Inflation

So, his guess at 2.5% real return would be something like (1.5% + 3% - 2%)? Isn't 3% a bit low for expected Dividend Growth Rate over the next 10 years?

[richard]

I don't see how anyone can reliably forecast returns. <snip>

You're just saying that because no one has.

Vanguard found the best predictor of ten year real returns is e/p (based either on normal p/e or Shiller's p/e10). Even so, p/e ratios have “explained” only about 40% of the time variation in net-of-inflation returns. Similarly, current yields on bonds are the best predictors of bond returns, best in the sense that everything else is worse, not best in the sense of very very good.

You should read any such forecast as having a high margin of error. Stocks at 5% plus or minus 10% might be reasonable for a ten year forecast.

[Tom_T]

Aren't we always saying that nobody can predict what the market will do this year? Ten-year predictions, even by a respected writer like Bill Bernstein, are not (or should not be) actionable, therefore they are nothing more than interesting reading.

[letsgobobby]

Reasonable to me, while recognizing he could be completely off. Naturally these are just midpoints of estimated ranges and even then the true values could fall outside of the ranges.

That said, in November 2013 I dumped 90% of my small/value US tilt and bought international with all of it, especially emerging markets, international small, and total international. In our 529s I bought DFA international value. I have not tracked how this decision has performed in the last 7 months - knowing my history, probably not well, as I tend to be a little early in these things. But my 60/40 portfolio is currently 60% international and 40% US and I'm very comfortable with this going forward.

[anil686]

I agree with the sentiment - we cannot predict future returns. If that happens, accumulators will be happy but it will be tough on those nearing or in retirement - JMO.

[ofcmetz]

I'm OK with my portfolio having a real yearly return of 2.5% to 3% after inflation. Actually, it is what I have been expecting. I think that it should motivate people to save more and spend a bit less. If we get better returns then we can adjust our plans accordingly along the way.

[bhsince87]

So that would imply a safe perpetual withdraw rate of 2.5%, right? And probably make a 30 year rate of 4% seem reasonable as well.

I'd be OK with that.

[cfs]

Internationals

Wow, based on the chart internationals are the place to be and things are not looking good for the Bond's Home Team. Good luck with your portfolios during the next ten years. Meanwhile, I will continue to navigate my ship based on my plan of intended movement, steady as she goes.

Thanks for reading.

[arcticpineapplecorp.]

I'm a fan of Dr. Bernstein, take his opinions seriously and have read most of what's he's written over the years.

That being said...since foreign developed stocks have the highest expected return of all the asset classes listed...why don't we all just put all our money there!!

Seriously, though...I guess I'm not concerned about the returns over the next 10 years since I'm not planning on using my investments for the next 27 years...and only then drawing it down slowly over the next 30 years after that. Can anyone tell me the returns of the asset classes between now and then? If not, I guess I'll just stick to the three fund portfolio and get what the market gives over time.

Thank you though for the information from Dr. Bernstein as I have not read this latest publication of his.

[Clearly_Irrational]

His stock return estimates seem overly pessimistic.

For a PE10 model you're looking at 1/PE10 which is 1/26.24 = 3.81% real

For an attribution model you're looking at (dividend yield + productivity growth rate + population growth rate) which is (1.81 + 2.2 + 1) = 4.02%

Expected inflation is (10yr treasuries - 10yr tips) = 2.61% - 0.38% = 2.23%

So my SWAG is 3.92% real or 6.15% nominal if you buy at current prices.

[Artsdoctor]

Yes. I am making my investment decisions approximating an average real return of 2.5% per year and I have been for at least the past couple of years. Last year was a very nice surprise, but with a 50/50 portfolio, I really have relatively low expectations. For me, this is fine because I have what I need and I'm more interested in preservation and slow growth than really anything else. If the average annual return turns out to be more, great; if it's less, I'll still be fine.

[MoonOrb]

I'm OK with my portfolio having a real yearly return of 2.5% to 3% after inflation. Actually, it is what I have been expecting. I think that it should motivate people to save more and spend a bit less. If we get better returns then we can adjust our plans accordingly along the way.

This is pretty much what I'm planning too so this doesn't really seem like bad news to me.

And the options are the same as always right? Reduce expenses, invest more, delay retirement, add risk. If my investments don't perform well enough to allow me to retire when I'd like, I'll probably focus on delaying retirement and reducing expenditures once in retirement; I'd rather put off the pain until then if I can, since there's a good chance I won't have to make those changes in the first place.

[berntson]

On an interesting side note, I was happy to find that Bill takes roughly the same approach to factor tilts that I have been taking in my own portfolio construction. He thinks that small, value, momentum, and profitability premiums will continue going forward, but that they will be less than half as large as they have been historically. This is built into his ten-year forecast.

Thanks to the limits of arbitrage, I think that we can still expect a boost from the premiums for small, value, momentum, and profitability, but not as much as in the pre-1992 era. Although tables 1-1 and 1-2 suggest perhaps a 2% premium for small stocks and 3% for value stocks, I would count on less than half of the previous premiums at best for these two factors going forward, as well as from momentum and profitability, for the reasons mentioned above. So figure on premiums of perhaps 0.7% for small stocks, 1% for value stocks, and thus 1.7% for small value stocks. Piggybacking off the 2.2% estimated return of the U.S. stock market, when calculated above for the upcoming decade, this 1.7% fillip gets us close to an expected real return of 4% for small value stocks, for example.

[IlliniDave]

Seems to be on the pessimistic side of reasonable to me, but still reasonable.

[berntson]

That said, in November 2013 I dumped 90% of my small/value US tilt and bought international with all of it, especially emerging markets, international small, and total international. In our 529s I bought DFA international value. I have not tracked how this decision has performed in the last 7 months - knowing my history, probably not well, as I tend to be a little early in these things. But my 60/40 portfolio is currently 60% international and 40% US and I'm very comfortable with this going forward.

Great minds think alike. My target domestic/international split is also 40/60. I'm most optimistic about international developed value. Given that Bill is working with the assumption of a 1% value premium for value funds, I suspect that his expected returns for international developed value are 6%. I also made room for more international by decreasing my allocation to domestic small value. I still have roughly 25% of my portfolio in domestic small value and microcap stocks.

[DVMResident]

Can Bill provide the 95% confidence interval these returns?

[investor1]

Wasn't he recommending to stay out of REITs last year? They are up >16% YTD.

Regardless of how much they are "up", they are still very opaque. The way the crash went down shows why that's a real problem.

[Leesbro63]

Is Bernstein saying 2.5% TOTAL real return for 10 years or average annual?

[givewell]

Must be annualized - or roughly average annual.

[209south]

I read chapter 1 today and love the book - as he says in the preface, it's relatively dense and will require some work. With equities and bonds at very strong valuations currently, I am planning for weak long-term returns - my personal plan incorporates real returns of 0% for cash and bullion, 0.25% for bonds/TIPs, and 3% for all equities/REITs - I am approaching retirement and my #s work with these estimates - I have a 35/65 portfolio so I hope his negative bond returns aren't realized!

[berntson]

Is Bernstein saying 2.5% TOTAL real return for 10 years or average annual?

They're definitely annualized returns. Sorry that wasn't clearer!

[MathWizard]

Even 10 years is too short. This numbers will still be dominated by the change in PE ratio.
30 year forecasts are a more reliable measure of the underlying (hopefully upward) trend

[JoMoney]

...What do you think of Bill's expected returns? How do you plan to deal with low expected returns?

Apparently, I have a lot more optimism then Dr. Bernstein. While I do think bond returns can be projected fairly closely, I give very little credence to predictions of stock returns and inflation.
"The expected never happens; it is the unexpected always"-John Maynard Keynes

How do I plan to deal with the markets returns? The same as what I'd do if it was expected to have historically high returns. I don't control the markets return, you can't push or pull a plant to make it grow. I'll focus on my job, growing my personal abilities, try to earn/save more. The stock market will do what it will do.... Now that I think about it, the "euphoria" of high future expected returns might make me a little more bearish on stocks. The bright side of high expected returns could make for higher current bond yields. Investors should demand a higher yield from bonds when stocks future "expected returns" are high. It seems almost unreal today to consider that one could have bought an investment grade bond fund yielding 7.5% back in 2000.

[Tycoon]

What do the gizzard squeezers say?

[Cut-Throat]

Does anyone know the returns for another decade that was worse than what he was forecasting ?

Since, we all plan for the worst case historical periods, a 2.5% real return makes me on easy street.

[deikel]

I have not read the book, neither would I claim to be as smart as the man, but if he means an average real return over the next 10 years of 2.5 %...

..that would leave room for another crash in say one year and a recovery over the next 9 years, right ?

That seems about right to me. We are still in a cyclical world where limited true gain is achieved until th enext real innovation comes along and as long as players are playing the market up and down....I will rebalance back and forth to make my average higher.

[berntson]

...What do you think of Bill's expected returns? How do you plan to deal with low expected returns?

Apparently, I have a lot more optimism then Dr. Bernstein. While I do think bond returns can be projected fairly closely, I give very little credence to predictions of stock returns and inflation.

I do wonder about this. Bill assumes that valuations will mean revert over time. Since valuations are unusually high, this could either happen because the market will have below average returns or because the market will have above average growth. Bill then goes on to assume that growth will be average, so concludes that the market will have below average returns. But it could be that the market reasonably expects growth to be higher than average. In which case, returns would be much better.

I don't know. I tend to think that markets tend to overestimate growth rather than underestimate it. But that doesn't mean that the markets are wrong this time.

[linenfort]

I'm glad it's impossible to predict returns. If someone had told me circa 2000 that the S&P would have a lackluster decade, a "nothing" decade, I wouldn't have had such a happy ten years, and I might not have been fully invested in time for the turnaround in the market.

But, I agree with others who said beating inflation by 2 or 3% is not terrible. Could be worse.

[dbr]

What do you think of Bill's expected returns? How do you plan to deal with low expected returns?

If you haven't placed an estimate of the range of error on your forecast around the numbers, then you really haven't said anything.

[berntson]

What do you think of Bill's expected returns? How do you plan to deal with low expected returns?

If you haven't placed an estimate of the range of error on your forecast around the numbers, then you really haven't said anything.

Bill uses standard deviation for his error ranges, though they are not included on the chart.

[JoMoney]

What do you think of Bill's expected returns? How do you plan to deal with low expected returns?

If you haven't placed an estimate of the range of error on your forecast around the numbers, then you really haven't said anything.

But even then, it's been shown that people tend to anchor around their initial guestimates and create an insufficient "range of error" around their rules of thumb used to create such predictions. Market returns don't necessarily follow a random walk or make neat distributions the way people like so they can attempt a range using mathematical probability scenarios.
The only certainty is uncertainty.

[asset_chaos]

I don't see the 2.5 from the numbers in the table and the postulated asset allocation of 40% bonds, 20% non-US (about 30% of the 60% sock allocation) and 40% US stocks. I get at best -1*0.4 + 5*0.2 + 2*0.4 = 1.4. Does Bernstein expect multiple expansion for the other 1 percent? I've ordered the book; next week I'll see the details for myself. This is but a preview.

[ray.james]

One of the issues with mean to reversion is mean value is slightly moving target as market get efficient and stability of the countries improve.

40 years ago...mean average of people would have been ready to take PE10 of 15 as an average. Today people are happy even with PE10 of 18 because of history of stability with USA, more number of investor penetration and market efficiency.

In the similar fashion, every decade passes acceptable risk of EM PE10 will rise. This will go hand in hand with the fact more countries jump from emerging to semideveloped to developed. EX: china, south korea, saudi arabia, Turkey, poland
Of course the risk of war/World war/severe natural calamities remain...which can reshape that risk, but in a global market its weight & impact on individual investor is decreasing.

Edit: Having said that, those returns still seem fine for any early accumulator, retirees. It is the accumulator in his second half of accumulation phase, that need to prepare for lower compounding.

[manwithnoname]

I just finished the first chapter of Bill Bernstein's new book. What a great read! Everyone should buy the book and read it. Now. Immediately. Don't finish reading this post. Buy the book and then come back. (No, Bill didn't pay me to say that. )

The first chapter ends with ten-year expected returns estimates for a number of asset classes. And they are not rosy. Take a 60/40 three-fund portfolio for example. Let's assume that the stock allocation is split 70/30 between domestic and international (this is roughly what Vanguard's target date funds do). The three funds are TSM, VXUS, and BND. Given Bill's estimates, a 60/40 three-fund portfolio can be expected to have a real return of around 2.5% over the next ten years.

What do you think of Bill's expected returns? How do you plan to deal with low expected returns?

This a joke right? no one can accurately forecast returns for the next 12 months, no less next 10 years. Its a waste of time to read stuff like this because it does not add any value.

Since I am not a neurologist I don't try to predict the future returns in the financial markets.

[berntson]

I don't see the 2.5 from the numbers in the table and the postulated asset allocation of 40% bonds, 20% non-US (about 30% of the 60% sock allocation) and 40% US stocks. I get at best -1*0.4 + 5*0.2 + 2*0.4 = 1.4. Does Bernstein expect multiple expansion for the other 1 percent? I've ordered the book; next week I'll see the details for myself. This is but a preview.

Sorry everyone. Bug in my spreadsheet. This isn't Bill's fault, since I was generating the expected return for a 60/40 portfolio from his chart. Given Bill's expected returns, the 60/40 portfolio should return 1.4% annualized over the next ten years, not 2.5%. I'll edit the original post to reflect this. Thanks asset_chaos!

I guess the next ten years look even more dismal.

[hoppy08520]

Just for kicks, and because I have too much time on my hands, I checked up on William Bernstein's estimated returns that he made back in 2002 in Four Pillars, using probably the same methodology as he did in the estimates he made in 2014 and which we are discussing in this thread. You can see his 2002 estimates here in the wiki:

http://www.bogleheads.org/wiki/Historic ... ed_returns

Summary
Overall, his estimates for the larger asset classes were fairly close to the actual results. His estimates in smaller asset classes, where you might expect more variance, were not as close.

Methodology
Using Morningstar, I tried to find Vanguard funds that hold the asset classes he estimated. Then I looked up their returns from 1/1/2002 - 12/31/2011, a ten year period. I also did this for the 10-year period one year before and one year after (Jan 2001 - Dec 2010 and Jan 2003 - Dec 2012 respectively) to try to introduce some "rolling returns" aspect to his estimates to minimize swings based too much on one set of start and end dates. We are talking about a 10-year estimate, remember. I then took all three sets of data to come up with a rolling CAGR.

Results
The results are shown below. You're also welcome to have a look at the spreadsheet.



The estimates for US stocks are fairly close to the actual results, as is the estimate for Large Foreign (I used a MSCI EAFE fund).

Where the estimates veered was with emerging market, foreign small value, REITs, and precious medal (who could have predicted the gold rush?), all of which performed far higher than the estimates.

Bonds also returned a bit higher than estimated.

My thoughts:

• Small asset classes (like international value, REIT) can have bigger swings and are harder to estimate.
• Back in 2002, emerging markets were also a smaller asset class. As we all know in hindsight, EM was the biggest boom asset class of the 2000's and clearly back in 2002, based on the conditions of the time, "the market" didn't see it yet.
• For the larger asset classes that typically make up the bulk of our equities (US Large Cap, Foreign Developed Large Cap), the methodology of the Gordon Equation that Mr. Bernstein used was fairly accurate. It was also fairly reliable for small cap stocks. Therefore, for those of you casting doubt on Mr. Bernstein's numbers for 2014, he wasn't that far off 12 years ago.

Disclaimers:

• I might not have picked the most representative funds. I'm sure the DFA boosters will suggest picking DFA's SCV instead of Vanguard.
• My methodology might not be sound.
• I might have made a typo transcribing these from Marketwatch to my spreadsheet.

[vitaflo]

What do you think of Bill's expected returns? How do you plan to deal with low expected returns?

Work longer? I mean really what else is there if everything has low returns? Everyone is basically in the same boat. I don't see how it would change my AA, and it's not like I'll suddenly pull out of the market based on a prediction.

[Clearly_Irrational]

What do you think of Bill's expected returns? How do you plan to deal with low expected returns?

Work longer? I mean really what else is there if everything has low returns? Everyone is basically in the same boat. I don't see how it would change my AA, and it's not like I'll suddenly pull out of the market based on a prediction.

In theory it could make you re-examine if financial assets are best choice, perhaps skewing the decision in favor of real estate or small business ownership for example.

[garlandwhizzer]

Interesting that 2 respected, astute, and influential figures in the investing world, Bernstein and Grantham of GMO, both have a very pessimistic outlook on US stocks and both expect negative real returns for bonds in the next decade. I personally don't believe than anyone can accurately predict the investing future even someone as distinguished as Bernstein. If these two guys are correct we should increase international equities but I'm skeptical that these forecasts are reliable enough to be actionable. Based purely on valuation differentials I shifted my equity portfolio to 2/3 US and 1/3 international at the beginning of the year, but I am hesitant to increase it from 1/3 to 1/2. Those investors who bailed out of equities in 2008-9 and stayed in cash or bonds through the this huge 5 year US bull market may have paid a dear price for their "safety" if Bernstein and Grantham are correct. The opportunity cost of missing out on this huge gain will be severe if bonds earn a real negative 1% and US stocks struggle into the low single digits for a decade. Hopefully this pessimism will be like the famous "Death of Equities" front page of popular periodical in 1982 (Time or Newsweek) that marked the exact beginning of the greatest secular bull market in US history, a bull that ran on for 17 years in both stocks and bonds.

Garland Whizzer

[protagonist]

With all due respect to Dr. Bernstein.....I have a lot of respect for him and what I have read of his work.....I would ignore anybody's financial predictions going ten years into the future.

If you have any doubts about this, go back ten years and read what the experts were predicting for today.

If Bernstein nails it, it is because he is lucky.
If Bernstein is way off, it is because he is unlucky.

Though whether he nails it or not will determine whether he will be a media star in 2024.

[hoppy08520]

With all due respect to Dr. Bernstein.....I have a lot of respect for him and what I have read of his work.....I would ignore anybody's financial predictions going ten years into the future.

If you have any doubts about this, go back ten years and read what the experts were predicting for today.

If Bernstein nails it, it is because he is lucky.
If Bernstein is way off, it is because he is unlucky.

Though whether he nails it or not will determine whether he will be a media star in 2024.

Actually, if you look about 5 posts up from your post, I tried to do this for his estimates back in 2002.

I don't think you should be so critical of his estimates. If you read the section in the Four Pillars where he outlines the methodology, there's actually a sound backing to it.

Someone remind me in 10 years to run the same numbers for his 2014 estimates (if I'm still alive).

[steve_14]

Interesting that 2 respected, astute, and influential figures in the investing world, Bernstein and Grantham of GMO, both have a very pessimistic outlook on US stocks and both expect negative real returns for bonds in the next decade.

Though their expectations are very, very different. For small cap value:

Bernstein: +4%/year
Grantham: -5%/year

So expect your small caps to either double or drop by half over the next few years . What's that Taylor says: "When experts disagree, neither likely has any idea". Well, something like that...