Executive Summary
For retirees who fear the impact of a market downturn on their spending, an increasingly popular strategy is just to hold several years of cash in a reserve account to accomplish near-term spending goals. As the logic goes, if there are years of spending money already available, the portfolio can avoid selling equities in a down market to raise the required cash, and clients don't have to sweat where their retirement income distributions will come from while waiting for the markets to recover.
Yet the mathematics of rebalancing reveals in the truth, even clients following a standard rebalancing strategy don't sell equities in down markets, rendering the cash reserve strategy potentially moot. On the other hand, some benefits still remain - although aside from an indirect short-term tactical bet, the most significant impact of a cash reserve strategy may be more mental than real.
Nonetheless, is the cash reserve bucket strategy still a viable option for retirees? Or is it just another bucket strategy mirage?
The inspiration for today’s blog post was a recent conversation I had with another planner regarding the use of so-called “cash reserve” strategies with retirees, where 2-3 years worth of cash is set aside in a separate account to fund the next several years of expenses. The concept is pretty straightforward: by having a portion of funds set aside to fund near term expenses, the retiree avoids the risk of potentially needing to sell stocks in a down market that might impair their future recovery.
The problem is, as I examined the approach further, I found that the cash reserve strategy – similar to other ‘bucket’ strategies I’ve written about previously – may be more of a mirage than a reality for protecting clients from selling out in severe downturns.
Rebalancing Portfolios In Downturns
Over the next year, the stock market experiences a severe 30% decline, causing the equity allocation to plummet from $500,000 down to only $350,000. In conjunction with the equity crash, an investor flight to safety causes bonds to rally, leading to a total return of 10% for the fixed portion of the portfolio. At the end of the year, the total portfolio is down to $900,000 (at $350,000 in equities and $550,000 in fixed). Absent any other cash flow needs, a rebalancing transaction would now occur. The investor would sell $100,000 of bonds to buy $100,000 of equities, to bring the $900,000 portfolio back to its 50/50 allocation.
However, the client does need $40,000 of cash for the next year. Yet observe the result that occurs. The portfolio was already selling bonds to buy equities; if the client needs $40,000 of spending money, the actual transaction that would occur would be to sell more fixed income – another $20,000. This would result in a total sale of $120,000 of fixed income (bringing the balance down to $430,000), of which $40,000 is spent by the client and the remaining $80,000 is reinvested into equities to bring the allocation up to $430,000, preserving the 50/50 balance on what is now an $860,000 portfolio (after the decline and after the withdrawal). The net result, simply put: in light of the rebalancing that was occurring anyway, the market decline actually meant the client sold no stocks at all, and in fact sold more bonds in order to help raise the cash needed to get the portfolio back in balance! Which means that what occurred already through rebalancing alone was exactly what the client would have wanted: the portfolio sold the bonds that were up and not only didn’t sell stocks that were down, but actually bought more of them!
| Equities | Bonds | |
| Starting Balance | $500,000 | $500,000 |
| After Market Returns | $350,000 | $550,000 |
| Trades (Rebal & W/draw) | +$80,000 | -$120,000 |
| Final Balance | $430,000 | $430,000 |
Of course, the reverse results would occur if stocks experienced the rally and bonds experienced the decline. Yet that, too, is exactly what the retired investor would want – equities get sold to raise spending cash when the stock market is up, and bonds get sold to raise spending cash when the stock market is down. The mathematics of rebalancing ensure that whatever asset class performed better carries the load for funding the withdrawals. (Editor’s note: It’s notable that mathematically, since the withdrawal of $40,000 is 8% of the $500,000 allocation to stocks, any time stocks and bonds perform within 8% of each other, a portion of the withdrawal will come from each category of stocks and bonds; nonetheless, the asset class with the greater return will still bear the majority of the withdrawal.)
As the math and the above chart reveals, it turns out that a cash reserve strategy is actually unnecessary to avoid selling stocks in a market downturn; the mere mathematics of rebalancing itself will ensure that the asset class with the better returns is sold first!
The Value Of Cash Reserve Bucket Strategies
Notwithstanding the mathematics of rebalancing, however, there are still two reasons why a cash reserve strategy may be helpful.
The first is that, as noted in the blog post earlier this month about bucket strategies, there may be some behavioral benefits to using a cash reserve strategy; simply put, clients often prefer to separate the purposes of money into various mental accounts anyway, and having a mental (and actual!) account for “the next 3 years of spending” may be reassuring to a client who is fearful in a market decline, knowing where spending will be coming from for the next couple years. As the math of rebalancing shows, this may not actually be necessary to protect the client from selling equities in a downturn just to generate the cash flows for withdrawals, but it could still be helpful to protect the client from the temptation of selling out of the markets entirely in a downturn when it’s clear that there will be a source of funds for spending needs while waiting for the recovery.
The second reason why a cash reserve strategy may be relevant is because it can actually function as a form of automatic market timing that makes short-term overweight or underweight bets on mean reversion to enhance portfolio returns. An example may help to illustrate.
Assume that our earlier client with a $1,000,000 portfolio has set aside $120,000 into a side account held in cash to fund the next 3 years of withdrawals (at $40,000 x 3 years = $120,000, ignoring any inflation adjustments); the remaining $880,000 is invested into a 50/50 portfolio, leading to an allocation of $440,000 in stocks and $440,000 in bonds. Notably, because of the cash held aside, the client’s total asset allocation is now 44/56, with $440,000 in stocks, and $120,000 + $440,000 = $560,000 in cash and bonds. Through the year, the market experiences a 30% decline in stocks, a 10% rally in bonds, and the client has a $40,000 withdrawal from the cash reserve, which is not replenished because the equity markets are down. As a result, the equities are down to $308,000, the bonds are up to $484,000, making the portfolio total worth $792,000, and the side account is spent down to $80,000. The portfolio is then rebalanced, bringing both stocks and bonds back to $396,000 each. On an aggregate basis, now, the client holds $396,000 in stocks, and $396,000 + $80,000 in bonds and cash, for a total allocation 45.4%/54.6%. The portfolio equity exposure has actually drifted slightly higher. If the bad market returns continue, where equities decline another -20% and bonds rally another 10%, and another $40,000 comes out of the cash reserves, the portfolio is now $376,200 in stocks, $376,200 in bonds, $40,000 remaining in the cash reserve account, and the allocation is 47.5%/52.5%.
| Equities | Bonds | Reserve Acct | Total | |
| Starting | $440,000 | $440,000 | $120,000 | $1,000,000 |
| After Market | $308,000 | $484,000 | $120,000 | $912,000 |
| Rebal & Withdrawal | $396,000 | $396,000 | $80,000 | $872,000 |
| Asset Allocation | 45.4% | 54.6% | ||
| Year 2 Market | $316,800 | $435,600 | $80,000 | $832,400 |
| Rebal & Withdrawal | $376,200 | $376,200 | $40,000 | $792,400 |
| Asset Allocation | 47.5% | 52.5% |
As a result of these dynamics, all else being equal, the cash reserve strategy actually ends out producing a more conservative asset allocation, but one that allows the equity exposure to drift higher in bear markets while the cash reserves are drawn down and the equities are allowed to run. If the market recovers and generates a positive rebound before the cash reserve bucket is depleted, the slight overweighting producing by the cash reserve strategy will create a small tailwind for the overall portfolio as the well-timed overweight to equities produces some value. However, there is little research currently available to clarify whether the small equity overweight timing strategy in bear markets (it was still only a 3.5% increase in equity exposure, even after a 2-year severe decline) offsets the fact that the portfolio has a lower long-term expected return by being underweighted to stocks in the first place due to the size of the cash reserves bucket itself. It is also not clear how much of a cash reserve is optimal to aid in these short-term asset allocation adjustments.
Ultimately, though, it's important to remember that, as I’ve written previously, the true disaster for either portfolio is a protracted period of merely mediocre market returns, which is dangerous not because stocks are being sold in a downturn – as the rebalancing math shows, they’re not! – but because ongoing withdrawals may deplete the total portfolio so far that there’s just not enough left invested to recover when the good returns finally arrive. And that’s a problem with either strategy, as even 3 years in cash doesn’t protect against a 5-10 year period of substandard returns – and notably, a decumulation plan can be destroyed by an extended period of low but positive returns as well, which the cash reserve strategy doesn’t protect against much better than a standard rebalancing strategy either.
So what do you think? Are cash reserve strategies just a mirage given the mathematics of rebalancing? Is the short-term tactical shifts they make a worthwhile source of alpha to pursue? Are the mental benefits for the client worthwhile, even if the strategy isn't actually necessary to defend against selling equities in a down market?
Michael,
Do you have evidence that stocks and bonds are negatively correlated? Your point makes perfect sense assuming stocks and bonds do not fall at the same time. I am curious to see the data for the historical correlations of the presented asset classes.
Alan,
Most studies I’ve seen peg the correlation near zero or slightly positive, but not necessarily negative.
It doesn’t really matter, though. Plug in -30% for stocks and -15% for bonds, and you still end out selling bonds to buy stocks and to raise the spending cash. Any time the returns are more than 8% different, in either direction, you sell the better performer to buy the worse one. (If the returns are close, you effectively sell a pro-rata portion of each, tilted towards the better performer.)
So the point still remains the same.
– Michael
Michael and Alan, I agree with both of you. While Michael’s point still is valid in a positive correlation scenario, Alan’s comment could be extrapolated to the extent that for several years returns are highly correlated. In this case, it would seem that funding all withdrawals with cash would have better results (assuming markets then go back to “normal”, whatever that means). One other way to look at it would be to hold “bonds” being the average duration, credit quality, etc. (i.e. diversified) plus US gov’t bonds for the “cash” component, due to their likelihood to continue to be zero/negatively correlated with stocks in a severe market such as we saw in 2008.
_Is_ the point the same? If both stocks and bonds lose money, as they do with -30% and -15%, you’re forced to sell assets at a loss. The cash reserve would prevent that, wouldn’t it?
Your argument assumes that if one part of the portfolio loses, another part will gain. That didn’t happen in 2008 unless you owned some treasury bonds. That’s what I worry about. In that case, the benefits of the reserve strategy don’t seem illusory.
Being forced to sell at a loss isn’t necessarily bad. There are plenty of Japanese and European investors that would love a second chance to sell at a 30% loss. You either have a plan or you don’t. You either have discipline or you don’t. Expecting all investments to just somehow work out by the virtue of patience is incompatible with reality.
I’m curious about one point: in down markets many clients won’t touch their principal, or at least scale back spending dramatically. So if a client only pulls out $20k instead of $40k, does it make a difference if that “bonus” $20k stayed in cash and was protected from down stock and bond markets?
Michael,
We’ve used the “cash bucket” concept with good success when judged primarily on the comfort level of our clients as measured by their ability to stick with their stated investment policy statement. The behavioral aspect, in my opinion, has great value.
I don’t see a mention of whether capital gains and dividends are being harvested or reinvested. By harvesting them to replenish the cash account, wouldn’t that have a significant impact on the strategy?
Michael,
If you are using re-balancing in a down market to raise cash, aren’t you robbing yourself (to a certain degree) of your opportunity to “buy low”? Or, in this case, “buy lower”? In the case that both equities and bonds are down, this could limit your ability to rebound, which was deterimental to many in 2009 and 2010.
I’m thinking that this may particularly impactful in more aggressive portfolios.
These bucket gimmicks are just a way of skewing your asset allocation in a more conservative direction. There’s no magic to it. If you’re pulling withdrawals from the cash reserve during a declining market, you’re making a tactical decision to reallocate toward riskier assets. Holding a larger cash bucket will always look smart during bad markets and less smart during ascending markets. As Mark said above, the greatest value of giving the cash allocation a special name is psychological.
Instead of using the cash bucket, what if you instead used a reverse mtg line of credit, especially for times when both bonds and stocks are down. This may also give you more time for your portfolio to ‘recover’ than a 3 yr cash reserve in the event of prolonged subpar market performance. Your thoughts on this, and the research done by Evensky?
Nope, that’s still an illusion. By drawing on a credit line you are shifting your asset allocation, just like drawing down your cash bucket does.
It also presumes that the market will recover before your 3 years of cash reserves runs out. If it doesn’t, then you *still* have to sell stocks to get your spending money — but instead of selling after 1 or 2 years of decline you are selling after 4 years of decline. Whereupon you probably wish you’d sold when you were 20%-30% down instead of the 40%-50% down that you’d now face.
Why would you have to sell after 3 or 4 yrs of decline? It’s a reverse mtg line of credit. Couldn’t you continue to draw until the market recovered? http://www.fpanet.org/journal/StandbyReverseMortgages/