Blog reader with the "Unkown" moniker brought up an important point about the weaknesses of using computer models to simulate the withdrawal of funds from a retirement portfolio. I cannot say that I fully understand the maths, but I think I can explain the problem to a lay person.
Most of the retirement simulations makes sense, but if you are unfortunately enough to face a nasty market crash just around a year or two after giving the middle finger to your corporate bosses, the odds of running out of money increases dramatically.
Initially I thought that there was no simple solution to this "Sequence of Return" problem - The foremost expert recommends a wonky approach called an equity glide path to resolve this issue. But a quick discussion with Kyith of Investment Moats gave me a useful insight to this issue that can be executed by laypeople.
Kyith summarised the challenge as one to survive the first recession after declaring early retirement.
So here's a possible fix to the problem - Create a second retirement portfolios just to deal with the first market crash after FIRE.
Suppose in the earlier scenario, our BBFA friend has accumulated $432,000 in the form of an equity portfolio to withdraw from social life completely and spend the rest of his days playing computer games whilst spending $1,500 a month. He may be wise to postpone retirement until he builds a second retirement portfolio that I call a "bear trap".
A bear trap allows the suspension of spending from the retirement portfolio until the market recovers upon the first market crash. Earlier on, I built a program to measure the duration of market crashes and the worst crash in the STI lasted 700 days from peak to trough. This means that a bear trap must be large enough to sustain our BBFA friend for around two years. Suppose he needs $1,500 a month, he will need a $36,000 portfolio to complete this bear trap.
Because this bear trap needs to account for a near term market crash, it should be placed in a cash deposit account that can give about 1.5% to 2% every year and not participate in any security that contains any market or credit risk.
This means that BBFA should pull the retirement trigger only if he has one $432,000 equity portfolio and $36,000 lying in a DBS Multiplier, OCBC 360, or UOB One account.
Suppose BBFA really encounters a market crash within 5 years of leaving the work force, he can simply take $1,500 from the bear trap account instead of drawing down his equity portfolio. This gives him ample time for his equity portfolio to recover from the nastiest downturn.
Of course very anxious readers will have concerns about the second and subsequent bear attacks. One possible solution is expand the equity portfolio slightly by 20% so that dividends can be directed to regenerate the bear trap after the first market crash. BBFA has around 6-10 years to fully regenerate the bear trap to prepare for the next downturn. This solution is not very attractive to me because it would be tantamount to reducing his withdrawal rate to around 3.5% instead of the usual 4.17% based on the model.
My inputs is that the bear trap can only trap one bear. the BBFA's main portfolio will have to grow after surviving this bear.
ReplyDeleteFor this to happen the market must move in a normal market cycle. And your average compounded rate of return when it recovers need to be more than the withdrawal rate.
The most simple solution is to use a lower withdrawal rate. This is a solution that not many willing to accept because they might really wish to give their boss the middle finger soon.
If not, a cash cushion like the bear trap guards you from the worse case scenario. But do note, there is no free lunch. the weakness of the cash cushion is that you lose opportunity cost. If the last few years of accumulation and the first few years of retirement is a bull, those cash have huge opportunity costs! You could have grown it to more, that you do not have to worry about some draw down.
Other than the cash cushion, the bond tent, is also similar. this one is better... because bonds are lower in volatility than stocks, but higher returns than cash. (i do wonder if the current situation is an exception) 3 years before retirement, shift your allocation mroe to bonds. then in the first 3 years of retirement, slowly shift to equity. You need adequate equity because if your early retirement is long, say 40-60 years, you need an excess rate of return over your withdrawal rate.
This was basically what I commented to Investment Moats a few months earlier, when he wrote his own article on portfolio divestment during retirement. :)
ReplyDeleteI heard it 1st from a US podcast where a financial advisor said that his SOP for the majority of his clientele, prior to retirement, would be to have a main portfolio plus a cash buildup of 4-8 years worth of living expenses. For US, their "cash" is more of short-term Treasuries and 3-month T-bills which pay at least the Fed Funds Rate.
8 years for the super kiasu, being the worst case in US for someone who retired just before the Great Depression, for his portfolio to recover back to breakeven with dividends reinvested. i.e. Oct 1929 to end-1937.
This solution addresses the problem of a bad bear market in the initial years. If a bear market occurs after 5, 7 years into retirement, then there shouldn't be an issue.
There were 1 or 2 academic studies I vaguely recall mentioning that the initial 3-4 years were the most critical for sequence of returns risk.
By avoiding drawdown of portfolio during the initial years if there's a bear market would shift the success rate up, as it gives time for recovery using the initial capital base, bringing your portfolio back to the median-to-upper-median zone of performance.
The point being not to avoid bear markets per se, but just to handle one IF it occurs in say the first 4 years of retirement.
Hi,
ReplyDeleteI maintain 84 months of expenses as a buffer for such circumstances. The majority of people may not agree with such approach. To each of our own.
Ben
That is actually what I'm doing at the moment. Have cash / fixed deposits / unvested company stock that can last me between 3 to 5 years, depending largely on how the spend patterns.
ReplyDeleteSo I have been taking out from this instead of drawing down from AUM.
One can adjust the spending in accordance to the dividend income amount. Flexibility is one of the way to stretch the stash longer.
ReplyDeleteBen