(1) Number of Years to Retirement
(2) Attitude Towards Loss
(3) Market Valuation Measured by P/E10
(4) Ability to Absorb Losses

While this sounds relatively standard, we measure the ability to absorb losses not only in investment terms, but in the analysis of each of our user’s personal financial condition. As an example, someone that does not have an emergency fund is directed to have a sufficient cash allocation to make up the difference. Someone who is 50 might have a 12 month cash reserve requirement versus a 3 month requirement for a 25 year old. This is one of the factors in determining the ‘ability to absorb losses’.

That’s the 50,000 ft. picture, but the most important part of asset allocation is understanding what you can afford to lose or to put it a better way, “What must I absolutely have available if something goes wrong?” Sometimes this is safeguarding against a major market decline, having enough money to carry through during times of unemployment, or having enough money outside of the stock market to not worry about what the market does. This last one is a personal favorite of mine. If you’re retired and have two years’ worth of cash and five years’ worth of bonds, that’s seven years to ride out a downturn in the stock market. When you have this much money outside of equities, are you more or less likely to react to market movements?

You have a solid blog and I’ll be adding an RSS feed into our site in Q1 2010.

Well, you’ve hit right on one of the trickiest questions in personal finance. There isn’t really a “best” answer to it. Over certain periods, history shows us that rebalancing annually would have been best. Over other periods, once every 3 years would have been best.

William Bernstein argues persuasively in his Four Pillars of Investing that rebalancing more than once per year is likely to be detrimental to returns due to the fact that bull and bear markets tend to last greater than one year.

Larry Swedroe argues in favor of doing quarterly checks using a “5%/25% rule.” According to this rule, you rebalance any time an asset class becomes out of whack by more than 5% of your total portfolio value, or by more than 25% of the value initially allocated to that asset class. (So, for example, his rule would suggest that you’re due for a rebalancing due to your bond allocation being more than 2.5% (or 25% of its initial 10%) off target.)

My own method is simply to use each monthly contribution to bring me closer in line to my target allocation, then do one big rebalancing each year.

My question is, how long should I wait to re-balance to my 90-10 again? Should I do this every year, say–on my birthday? Or every 3 years? Quarterly? Any advice??

-Grandpa

That’s accurate.

But looks at the returns from July 1921 through 1929. They’re fantastic. They make it worth having gone with a high stock allocation.

Plug a P/E10 of 15 into The Retirement Risk Evaluator and you get a safe withdrawal rate for an 80 percent portfolio of 5 percent. The SWR is calculated assuming that you will see the worst 30-year returns sequence we have even seen in U.S. history. Assuming the worst, a 5 percent real takeout worked with an 80 percent allocation. That’s not at all bad.

Now try some lower stock allocations. What do you see? The SWR does not go up. There’s nothing wrong with an 80 percent stock allocation for a retiree at that price level.

Now. I think it’s fine to go with a lower stock allocation. At 50 percent stocks, you still have an SWR of 4.8 percent. You are going with less stocks and not giving up much in the way of SWR. That makes sense.

But I think it is fair to say that the big danger to stock investors is overvaluation, not retirement. We are warning people about the chance of getting stung by a jellyfish and doing nothing about protecting them from man-eating sharks.

Rob

Not necessarily. As nfs, a poster at the Retire Early Home Page pointed out ;

“In July 1911, PE10 was 15.08, a moderate valuation level according to everything you have ever written.

The real value of the S&P 500 price was 206.45. Ten years later in July 1921, the real value of the S&P 500 price was 69.98, a 66% decrease in real terms.”

It would seem that the retiree might be well advised to consider some rendition of the old rule-of-thumb regarding age in bonds. And others as well with little tolerance for the inherent volatility of equities.

I don’t have the words in front of me but my recollection is that Greham suggested that stock investors might want to go with a stock allocation of 75 percent at times of low valuations, 50 percent at times of moderate valuations and 25 percent at times of high valuations.

Graham is the person who came up with the P/E10 concept that is now used by Robert Shiller and John Walter Russell and others well informed about the effect of valuations on long-term returns. P/E10 is different from P/E1 (the most commonly cited valuation metric) in that it smooths out the earnings numbers (P/E10 is the price of the index over the average of the last 10 years of earnings). One of the reasons why there is so much confusion on valuation topics is that too many people test the effect of valuation using P/E1 and P/E1 is unreliable. The problem with P/E1 is that the earnings number is overstated during economic good times and understated during economic bad times. You want a smoothed-out number that reflects the overall economic reality that will be applying in the future.

The fair-value P/E10 number is 14. 7 is extremely low. 21 is dangerous. 25 is insane. We were at 25 or above for pretty much the entire time-period from 1996 through 2008. We are today at 16, which is fine. The long-term (10 years out) value proposition for stocks is today strong.

Rob

Sounds sensible. Did Ben Graham suggest a rule of thumb? I think we can be safe in assuming that “normal” p/e ratios should not change markedly – there’s no such thing as a “paradigm shift” in the financial markets.