PE10 and the 4% SWR

[2B]

There was a recent thread (I won't resurrect) that discussed a recent article claiming that the PE10 during the initial phase of retirement determines the actual 95% SWR. The premise was low PE's at retirement allow a higher SWR than high PE's.

John Greaney reviewed this data and has an interesting "rebuttal" on his website (see link).

Another Look at Safe Withdrawal Rates and PE Ratios

As you can see from the first plot, there is statistical justification for the higher SWR at low PE's but the standard deviation is so large as to make the correlation almost meaningless. "On average" the thesis is supported but there are wide ranges of 95% SWR at every PE10. Unfortunately, the risk of failure is still substantial for many low PE10s with a SWR over 4%.

Another important reason to always look behind the data before jumping to a hasty conclusion.

"Figures don't lie but liar's figure." "Correlation does not determine causation." These are two important things to remember when statistics are thrown at you.

[REWahoo]

Quote:

Originally Posted by 2B "Figures don't lie but liar's figure." "Correlation does not determine causation." These are two important things to remember when statistics are thrown at you.

You left out the third 'important thing': "Numbers are like women - once you get them down you can do amazing things with them."

[2B]

One reply and already hijacked!

[aenlighten]

Looking at the historical data, there have been times when higher withdrawal rates were justified and these were when the market was very cheap, often after a decade of poor performance. The problem is these episodes are rather rare and most of the time the market has been more or less fairly valued and offered only a 4% safe withdrawal rate. The biggest problem with PE10 is it is backward looking so while the market was overvalued in 2000, that may no longer be case. That doesn't mean it isn't going to get cheaper, but 10 years may be more significant from a human memory standpoint than an intrinsic value standpoint.

[REWahoo]

Quote:

Originally Posted by 2B One reply and already hijacked!

Hijacked? On the contrary, I simply added a supporting statement.

[haha]

We could have quite a lot of fun with the topic of numbers and women, but today I am not interested in that stuff, at least not for another hour or two. So here is my post on the Greaney study referenced.






What comes barreling out at me from this graph is that at least under the groupings in this study stock allocations above 60% do very little good, even when starting PE10 is quite low. And conversely, allocations above 60% can do a fair amount of harm with the higher initial PE10 values. It looks like under most circumstance you want in the neighborhood of 40 to 60% stocks. 60% looks better, but not by much and it will certainly be more volatile so there is room here for individual preferences. Since we never know what part of a distribution we are drawing from, conservatism might favor shading down toward 50% unless PE10 is quite low.

If I were using this it would not be so much to bump up the SWR above 4% , but rather to give me better overall knowledge about the drivers of success.
Ha

[2B]

As we all see, past returns are variable. As we all know, past returns are not indicative of future returns. This leaves us with the necessity to plan but to remain flexible.

I'm still going with my two tier plan. I've got basic living expenses covered for 20+ years with fixed income/pension/SS. Travel and extra luxuries will come out of the equity bucket after replenishing the fixed bucket.