In an earlier post on asset allocation I talked about the notion of a 50/50 split of your assets between stocks and bonds at retirement.  I also mentioned the “4% rule” for calculating withdrawals from your accounts in retirement as a good number.  Where did these come from?  It’s not simply a compromise between the assets, and we can show why it makes sense.  With math!  Hurray!

Previously I showed how we can use Monte Carlo analysis to predict probabilities of certain events, by running many possible future outcomes using prior years’ mean (average return) and standard deviation (variation of a given year’s performance from the mean).  In the case of retirement, the goal is for the money to last for the rest of your life.  Let’s assume you are taking good care of yourself and you are going to have a nice, long 30 year retirement.  And because of inflation, you’re planning on increasing your withdrawals by 3% every year.  We can run Monte Carlo analysis for different asset allocations to see what is the probability you will still have some money left after living off your nest egg for these 30 years.  I did a simple analysis with only 100 cases just for illustration, most Monte Carlo analyses would use many more, but this should be reasonably close.  Also, here I’ve assumed that stock and bond means and standard deviations will mimic those of the past.  Again, not a guarantee.  Here’s what I got:

Probability of your money lasting if 100% in stocks: 83%

Probability of your money lasting if 100% in bonds: 39%

Probability of your money lasting if 50/50: 90%

These results should look weird.  So all stocks is pretty good, all bonds not so much, but if I allocate some of my stock money to these (worse performing) bonds my results get better?  Huh?  Yep.  Remember your goal for this money: last 30 years.  It is not maximize the most money I can possibly make, that would be in stocks.  What happens is the addition of the lower variable bonds help smooth out those cases where stocks get hammered, and increase your chances that your money will last.

Is 50%/50% perfect?  Not really.  Here’s my analysis for more stock bond ratios, for convenience graphed as % of assets in bonds:

This analysis shows an ideal is actually about 60% bonds, but what’s clear here is you need the growth power of stocks to help your returns, but there isn’t a lot of difference between 40%/60% to 70%/30% (bonds/stocks) in terms of probability of the money lasting.  I like to lean a little heavier toward stocks with bond yields so low lately, but I think anywhere in this range is fine in retirement.  The other thing to notice is, how comfortable are you with a 10% chance your money will run out before 30 years?  If not, the 4% rule might be more of a 3.5% rule.  Or, as I mentioned earlier, you can throttle back your spending in the years of stock/bond downturns.  Living to your means, always a good idea.