Friday, April 27, 2012

William Bernstein's "The Retirement Calculator from Hell"

I'm working on several different projects now, and while nothing is finished enough to post yet, I'm hoping to describe lots of interesting results in the coming weeks. 
 
A couple minor notes... Congratulations to Michael Kitces, who joins a rather elite list of individuals in Investment Advisor's Top 25 list for most influential people for advisors in 2012 and beyond. 

Doug Nordman wrote a very interesting piece, "How much will military veterans leave on the table?" which explores some of the themes I've been describing here recently.

He also mentions William Bernstein's classic 5 part "The Retirement Calculator from Hell" series, which was written around 10 years ago. I finally read all of these, and they are quite worthwhile. A few comments on each:


Part I is no surprise to people who've been following my blog, but it certainly was big news when he wrote it. In 1998, people were rediscovering the notion that constant inflation-adjusted withdrawals above 4% were prone to wealth depletion much faster than expected. He shows examples of this. It's all about Bill Bengen's sequence of returns risk.
  
He describes about using Monte Carlo simulation as an alternative to historical simulations. Also, importantly, he suggests building in forward-looking assumptions to the simulations to account for the current return environment. He suggests using lower volatility for stocks than seen historically to account for the mean reversion in stock prices that is otherwise missing from the simulations. That is interesting, but after thinking about, I think maybe this isn't such a great idea again because of sequence of returns risk. Even with longer-term mean reversion, the impacts of what happens in the early retirement period have the biggest impacts, and there you want to make sure your stock returns have enough volatility to show the full range of possibilities.
 
Maybe this is the most famous part of the series. Here he suggests forgetting about success rates above 80% when thinking about 30-40 retirement periods. Your simulations show you about the investment risk, but they ignore the various possible political, economic, and military failures that can happen over such a long period. This is a very important point.

I also like this quote, which I think ties in well with the idea that minimizing failure rates is not necessarily the most important objective for retirees:

The historically na├»ve investor (or academic) might consider reducing his monthly withdrawals to a very low level to maximize his chances of success. But history teaches us that depriving ourselves to boost our 40-year success probability much beyond 80% is a fool’s errand, since all you are doing is increasing the probability of failure for political, economic, and military reasons relative to the failure of banal financial planning.


This one is about demography. Raising the retirement age is the only practical solution the aging population, because even if everyone saves a lot, there is still a limited number of younger people producing goods and services for the older people to consume. Also, what matters is not just the absolute amount of your savings, but rather how much you saved relative to everyone else. Important food for thought.

This is about how happiness is related to where you stand relative to your peers, not in absolute terms. As standards of living increase, this means saving even more just to stay at the same level of happiness. If real per-capita GDP grows at 2%, then a 2% investment return is really more like a 0% return in terms of your relative position in society.



5 comments:

  1. I'm curios about your thoughts on:

    "what matters is not how much you save, but how much more than everyone else you save. In a world where everyone saves as if they’re going to retire at fifty-five, or even at sixty-five, none can."

    Surely there can be an AMERICA where everyone saves, right?

    ReplyDelete
    Replies
    1. Surely everyone can save more, but he is arguing that there are limits to how effective such savings will be.

      This is a point getting at the real economy of goods and services, rather than just paper assets. If baby boomers suddenly save a lot more and own more stocks and bonds and then retire, we have a fundamental problem that there may not be enough workers left to provide the goods and services that the retirees want. These shortages will lead to inflation as retirees bid up prices, which reduces the value of retiree assets. He is getting at a practical problem of the role of demographics and what happens when there are few workers per retiree.

      We could get around that with more immigration, or if productivity improvements allow for more output per worker, but that is the general point he was making.

      But beyond that, in an America where most people do not save, there can be great benefit to the few who do. They can enjoy a nice retirement.

      Delete
  2. It's interesting to see how these two withdrawal strategies perform under these measures. I've found dealing with utility of consumption particularly challenging and have limited myself to utility of bequests--positive or negative (running out of money). One perhaps technical point--I wonder if the kinked utility function I used for bequests should be applied to consumption. I liked the kinked function for bequests because there seemed to be such a fundamental change between running out of money and not running out of money. (Negative bequests are different in character than positive bequests.) With consumption, wouldn't there be just a smooth concave curve--every time you give up something and consume less, you give up something more valuable. Also, re: utility of consumption, Wade makes point that the change from the prior year may be much more important in a utility sense than the absolute level of consumption. I haven't studied the utlity of consumption that much, but I wonder if that's a weakness in the economic studies that apply utility functions to absolute levels.

    ReplyDelete
    Replies
    1. Joe,

      Thanks. Perhaps I doesn't make a lot of sense to use the kinked curve at the floor level when looking at utility of spending, but I think this matter is of much less importance than the general issues about the shape of the curve with regard to how quickly utility plummets as you fall before the floor level. Both approaches can give different degrees of plummeting, and this is where the hard decisions must be made.

      About the floor being linked to the previous years spending rather than an absolute floor, there have been theoretical studies of this now using the idea of habit formation. Mike Zwecher reviews some of these in his Retirement Portfolios book. Joe, I think that is a pretty worthwhile book for you.

      Delete
  3. Well actually as one day or another everyone has to be retired and for that majority has to do some planning for that. According to me saving is not the only option to make their lives better after retirement, but rather than this, investing in the right place matters a lot too. So if one has started planning for the future then they should explore the world and invest their money at the right place which can benefit them for the very long time.

    ReplyDelete