A fellow advisor wrote in a professional online forum last year, “[I]n rocky times I turn off dividend reinvestment and let the cash build so if they need money from me, we don’t have to sell principal.”
I want to comment not only on that, but also on so-called bucket strategies that I don’t think I’ve written up anywhere.
This strategy doesn’t make any sense. (Mathematically, that is. It may feel good emotionally to “do something” rather than just strategically rebalance.)
When a company pays a dividend, the stock price goes down by the amount of the dividend. You are essentially selling the stock. Imagine a company that paid no dividends, and you sold an amount equal to their earnings yield. That is equivalent (mathematically) to not reinvesting the dividends of a company with a 100% payout ratio. But I don’t think anyone would think selling some shares of AMZN or BRK isn’t selling.
Now I suspect what was meant was at the asset class (not company) level. But still, why would changing the asset allocation (from stocks to cash) by an amount that happens to be equal to the dividend yield be optimal? When dividend yields are 1.3% (2021) that’s the right amount to go to cash (annually) but when they are 7.4% (1950) that’s the right number?
Are we avoiding invading principal in real or nominal terms? So today you need to reinvest everything (plus some) to maintain a real principal balance or in the 1970’s you can spend all your “income” even though the purchasing power of the portfolio is plummeting? Both seem silly, but you get either one silly outcome or the other if you follow these types of rules. The 4% rule (and most variations) and MPT are based on expected returns and risk. There is no place for “but the return comes from dividends not appreciation so it’s different.” It’s not different. Total return matters, not yield (except as a component of the total return).
In addition, “in rocky times” probably means the market is already down so you are (effectively) selling stocks lower. (Unless this is an actual VIX strategy that actively allocates based on the VIX and goes more to cash when stocks spike up as well as down.)
This is actually just disguised market timing because you have to make a subjective call on “rocky” and “not rocky” – not only to stop/start dividend reinvestment, but also to determine when to redeploy the cash that has accumulated.
Similarly, some advisors (including some I respect) employ a “bucket strategy” that divides the client’s funds into short-term and long-term buckets (and sometimes into intermediate-term too). While I’m sure that has psychological benefit, I don’t think it works the way people think. The idea (as I understand it) is that the client has some amount in cash to cover the next 2-5 years of expenses, so they don’t have to sell if/when the market is down.
First, that is market timing (again) because it makes the decision of when to replenish the cash account subjective (it implies you can tell when the market is “too low” to replenish the cash). If you always keep X years in cash then it’s just an allocation with a fixed dollar amount in cash – which seems odd, and I don’t think that has any empirical support in the literature as an optimal allocation strategy.
Second, it doesn’t work that way! In any significant downturn you would be buying stocks even after taking a draw, not selling them. Here’s the math:
Suppose we have a client who is 60/40 stocks/bonds with a 4% initial draw. Assume the market value of the 40% doesn’t change. How much would the 60% have to decline before we are buying stocks anyway? 10%. Let’s use dollars. On a $1 million portfolio, the $600k declines to $540k. $40k is withdrawn from bonds (you are overweight bonds now) for the client’s living expenses, leaving $360k. Total portfolio value is $900k. $540k/$900k is 60% and (obviously) $360k/$900k is 40%. Any decline larger than 10% in stocks would lead to buying stocks to rebalance back to target.
In short, if you rebalance, you can skip the bucket strategy as it makes no sense (mathematically) unless the investor has a pretty high equity allocation, or a pretty high withdrawal rate. I think people just say this because it sounds smart, but they never actually thought about it and certainly didn’t do any math!
At a 20% stock underperformance relative to bonds, you would need an 8% withdrawal rate (off the original values, which would be over 9% of the current values) before you were selling stocks. Again, the worse the bear market, the less likely you are to be selling stocks. If there is a 50% stock decline you would need a 20% withdrawal rate policy to be selling stocks (again off the original value, it would be almost 29% of the value after the drop).
But what if you were 80/20 to begin with rather than 60/40? Then on a 20% downturn you would take anything over a 4% withdrawal from the stocks (again of the original values, I’m using the “4% rule” methodology where you don’t change the withdrawal due to market returns, but I’m ignoring inflation for simplicity of exposition) and you would be “selling” in a bear market. If we are 80/20 and had a 50% market underperformance you would need a 10% withdrawal rate (pre-drop, almost 17% of the new value) before you sold any stocks.
In the most catastrophic downturns, you are never going to be selling stocks – so I have no idea what a bucket strategy is supposed to do, but I’m pretty sure it doesn’t do it.