Wednesday, June 20, 2012

Review of William Bernstein’s new e-book on Lifecycle Investing

William Bernstein is beginning an experiment of sorts. He is writing a series of low-cost e-books called Investing for Adults. The first of these is now available: The Ages of the Investor: A Critical Look at Life-cycle Investing (Investing for Adults).

As he notes, these books are not for beginners. They are for adults who have outgrown their beliefs in stock-picking fairies, market-timing fairies, and risk fairies. I think the first two are clear. The risk fairy tells you that the risks of holding stocks decline with time. Despite the name “expected returns,” you cannot simply expect greater returns for having loaded more funds into volatile assets.

The book is relatively short and is fully of interesting insights. I think it is definitely worthwhile. He divides investing life into three phases: the beginning, middle, and end.

The Beginning

For the beginning phase, the overwhelming consideration is that one’s human capital (the present value of future earnings) will likely dwarf in size one’s portfolio of stocks and bonds. This is a justification for allocating more assets to stocks when young. As well, he discusses in detail the interesting issue of lump-sum investing versus period contributions. Young people are generally forced to contribute new funds as they age (they can’t borrow their human capital at the start) and this reduces volatility and allows them to take advantage of market dips.

Bernstein discusses age-in-bonds, the Ayres/Nalesbuff strategy of leveraging to 2:1 in stocks when young in order to better balance stock holdings with human capital, call options on a stock index as another way to leverage assets, and also the Fama-French approach to focusing on risk factors related to small-capitalization and value stocks.

Despite the fact that young people can take much greater financial risk, Bernstein argues that young people are actually quite risk averse and should probably start with no more than 50% stocks. Then, they should gauge their reaction after experiencing their first big market drop to decide whether their appropriate stock allocation might be more or less than 50%.

The End

As for the retirement phase, Bernstein describes the goals-based approach to retirement income flooring. The objective becomes to secure lifestyle needs and take as little risk as possible. Retirees should build two separate portfolios in retirement: a “liability matching portfolio” (LMP) structured to support desired lifestyle spending with inflation-protected and secure assets, and a “risk portfolio” with any remaining assets which can be used for luxuries and bequests. As RIIA says: first build a floor, then expose to upside.

He describes various options for the LMP portfolio, including inflation-adjusted single premium immediate annuities, a ladder of TIPS, delaying Social Security to 70, and a mix of TIPS and deferred annuities. He also discusses part-time work and a portfolio of stocks and bonds. He notes that about half of the dividend yield on stock holdings could be treated as part of the LMP.

He indicates that it is tough to know which risk is greater: running out of funds because you live longer than the end date for your bond ladder, or running the risk of a systematic financial crisis that wipes out the ability of the insurance company and the state guarantor to provide you with your annuities payments. 

He does look at several examples of people with different spending needs. A rule of thumb he provides is that by age 70, people should have enough safe assets to fund at least 20 years of spending needs after accounting for Social Security and other pensions. Of course that is very tough to do, especially in today’s low interest rate environment.

The Middle

This period is saved for the end because it is about shifting from the aggressive early part of life to the defensive latter part of life. There’s lots of interesting tidbits here, including the idea of waterfalls which get at the heart of intergenerational risk. Those born just a few years apart may have very different lifetime investing experiences. To me, this suggests a need for intergenerational risk sharing as we enjoy with programs such as Social Security. Relying solely on defined-contribution pensions means relying much more on luck.

All in all, this is a very worthwhile e-book and I am looking forward to the next volumes in this series.


  1. A remark on the paragraph “He indicates that it is tough to know which risk is greater: running out of funds because you live longer than the end date for your bond ladder, or running the risk of a systematic financial crisis that wipes out the ability of the insurance company and the state guarantor to provide you with your annuities payments. “
    I don’t agree, at least as to immediate fixed or inflation-adjusted annuities. We have just gone through a particularly difficult economic period. To the best of my knowledge, not one annuity provider has defaulted. A few have been downgraded, but none to the point of danger. Indeed, there are few cases where an annuity provider has ever defaulted in the history of the industry. There are of course no guarantees that the future will be like the past, but so far so good.
    In addition, consider the economics. When you buy an immediate annuity, the insurance company basically builds a bond ladder to provide the payments because it is willing to bear mortality risk but not interest rate risk. (True on a group basis, if not individually.) So if for some reason many annuity providers fail (“many” because the state guarantors can handle isolated failures), why should a bond ladder be safe? The assumption seem to be that individuals can do a better job selecting bonds than can the annuity providers – good luck with that, and don’t take advice from con men or from incompetent advisors or from relatives out to feather their own nests. In fact, the only situation likely to result in the failure of annuity providers but not bond ladders is a sudden lengthening of life spans caused by a drop in mortality rates (that’s the risk the annuity provider retains) and, if that happens (seems very unlikely but not impossible), that bond ladder won’t be long enough for many retirees because they will outlive it.
    So it seems to me that the advantages point toward the annuity.

    1. Thanks for sharing. I've heard this same sort of point made by people I trust who have experienced working at insurance companies. The argument is basically that the investments behind annuities are not really at risk for blowing up even with a bigger crisis. Or at least, the event that might trigger it, such as a massive default by the US government on its bonds, would also destroy any possible retirement strategy. So some folks may be exaggerating the potential risks of insurance companies not being able to make their annuity payments.

    2. The GAO looked an insurance company failures in the early 1990's, after Executive Life failed. IIRC, some annuitants got 70% of their payments for a year and then began receiving 100% of their payments.

    3. Boy do I disagree. If insurance companies stuck to writing SPIA than I'd agree that annuities were a lower risk than DIY bond ladder. However insurance companies don't just write life insurance policies and SPIA, they also write annuities with a COLA provisions, Equity Index Annuities, Long Term Care Insurance and host of other products which can be mispriced. The also are subject to making bad investment choices. You don't have to be at all paranoid to suspect that a portfolio heavily weighted with 30 Year Tbonds with <3% yield, may have difficulty providing a COLA annuity written 5 years ago.
      Finally, insurance companies are run by people some of whom are incompetent, greedy and dishonest.
      After reading the classic book "Popular Delusions and the Madness of Crowds", it seems inevitable that there will be future insurance industry crisis just like there will be future banking crisis..
      I'm with Bernstein tough to figure out which risk is worse.

    4. Hi,
      Thanks for sharing. That's a powerful quote. I don't really know enough about what goes on inside of insurance companies to add much to what you've said. Thanks again.

    5. This is getting a bit circular, but here is a link that goes a bit deeper counting actual insurance failures over time. The article actually references this article.

  2. I agree with everything Anonymous says and he says is very well. Re: the risk of a dramatic increase in longevity, there is a natural hedge there because the vast majority of insurers have substantial blocks of inforce life insurance that would benefit.

    1. That is a good point too! Do insurance companies consciously try to optimize their ratios of annuities and life insurance sales so that they will be hedged very well for surprising changes in mortality rates?

    2. That seems like a rational thing to do Wade. However, TIAA, a huge company, does not do this. IIRC, virtually all of its liabilities are deferred or immediate life annuities.

    3. Note that TIAAs annuities are participating annuities; the company does not guarantee the size of the payment. So it would be able to adjust payments if necessary. That's quite ununsual.

    4. I should also have noted that TIAA is an annuity company that offers life insurance. Most other companies in this space are life insurance companies that offer annuities. One might expect differences in balance between TIAA and the other companies.

  3. It was reassuring to read that Bernstein thinks a 2% withdrawal rate is bulletproof at any age below 65.

  4. Unfortunately, Bill doesn't take the asset/liability matching framework to its logical conclusion: one can also target a minimum income in retirement before retirement by buying deferred life annuities (or at least simulating them by buying long term TIPS in the accumulation phase and then using the proceeds to buy life annuities at the beginning of retirement). I think DFA's managed DC works in this way.

  5. Thanks for the comments, everyone.

    Alec, yes, you are describing DFA's Managed DC pension. I wrote about it here:

  6. Wade, met you briefly at the RIIA spring conference and have been a semi-regular reader since. Really enjoy your work.

    Your summary of this book includes the sentence
    "Young people are generally forced to contribute new funds as they age (they can’t borrow their human capital at the start) and this reduces volatility and allows them to take advantage of market dips."

    I have seen this belief expressed in other articles as well, and it makes sense to me. For instance, the Ayers/Nalebuff book "Lifecycle Investing" discusses the case of the young investor who loads up on LEAPS.

    My question is: should volatility be considered beneficial in and of itself to a young retirement saver? Let's say you can't invest in LEAPS, but you can invest in a very volatile portfolio of equities. In the extreme, might it make sense to invest in a portfolio loaded with idiosyncratic security risk simply because it is more volatile? How far should we take this notion of taking advantage of market dips, which might be called dollar cost averaging?

    1. Dan,

      Thank you. For some bizarre reason, your comment was moved to spam. Sorry about that. I'm trying to remember you, and I think you were with Sam at lunch with me and Bob. Is that right?

      It's a good question. I think the idea is that a young person can basically afford to lose it all in the stock market, because their human capital is so much larger than their financial capital and they will still have plenty of time to recover their losses.

      However, something Bernstein emphasizes is what the psychological impact on the young person will be. If losing it all scares the person from ever investing again in the stock market afterward, then it is very destructive and a bad idea.

      So I think the answer has to be that it depends a lot on the person. Bernstein actually suggests that young people should not be more than 50/50 until they experience their first market drop and then gauge their reaction. But for those who can stomach the volatility and have the flexibility to recover, this is a strategy that _could_ pay off.

    2. You have a good memory :-)

      Thanks. The 50/50 thing is a decent idea if you are unsure. The new NEST program in the UK takes a similar approach with their default asset allocation strategy.

  7. Wade-

    I've started a thread over on about WB's new eBook. One of the tougher questions is how, in today's low % rate environment, does one create the LMP (for essential expenses) described by Bernstein?

    I'd be interested in your thoughts.

    1. Hi, I agree that it is tough. It's a bad time to retire. I'm not really sure what to do. One could take on a bit more risk to get higher yielding corporate bonds. I to try using a stock fund with higher dividend payments. Or just go ahead and buy the annuity now, as waiting for rates to go up may not work out when you are drawing down principal in the mean time.