8.5 Timing the market

The prior series 8 articles on “Investing over time” covered investing plans for both young and “old” investors.  A mindful approach to investing advocates buying and holding mostly low-cost and reasonably diversified index stock funds as soon as long-term money is available for investing.  The only time you should sell is when you need that money for its planned investment purpose (like retirement) during the so called “withdrawal” phase.

The whole idea of buying and selling stocks (or any other investments) over time is antithetical to the mindful approach.  Apparently a mindful investor need not even consider the concept of “timing the market”, which involves attempts to sell when stocks are “high” and buy when stocks are “low”.  So, this article is over before it began?  Well, not quite.

Given that a mindful investing approach requires an empirical and logical reason for doing anything, we need to cover two concepts about timing the market, which can be distinguished by their relative time frames:

  • Short-term timing – This means buying and selling stocks in time spans roughly less than a year (daily, weekly, monthly).
  • Long-term timing – This means investing decisions based on stock market movements that occur over many years, such as the time between when you start to invest and when you withdraw money in retirement.

So, let’s tackle each concept in turn.

Short-term timing

Short-term timing is usually what people call “timing the market”, whether they recognize it or not.  We need to be sure exactly why we wouldn’t attempt to time the market on daily to monthly time scales.  All evidence indicates that timing the market on a short-term basis is impossible.  We’ve already covered much of the evidence we need to easily reach this conclusion.  Here’s a summary from past articles:

  • Article 4.1 covered that individual investors consistently under perform the funds they invest in.  Most of this is due to buying and selling behavior, instead of just holding on to the funds already purchased.
  • Article 5.2 presented ample evidence that almost no investors regularly beat market indices through “active” management. This includes individual investors, professional managers of active funds, and hedge funds.  Further, it’s at least questionable whether even the very best investors like Warren Buffet are mostly clever or mostly lucky.  Active management includes the timing decisions of when to buy and sell.
  • Article 5.2 also presented that attempting to time the market during volatile periods can result in missing out on big gains from a few key days.  For example, if you missed just the biggest 30 up days in a 28-year period of S&P Index investing discussed in Article 5.2, your overall return would go from 11.1% to just 5.5%.  Your returns would be cut in half.
  • Article 5.3 showed that trying to pick the few very best “active” managers (if they even exist) is probably even harder than picking and timing the best investments yourself.  So, attempting to time the market through a professional proxy is equally infeasible.  More importantly, such active timing is not necessary for most people to achieve their investing goals.  So, there’s little purpose in attempting to perform this difficult feat.

If you’ve read some of the content on this site, you’re probably asking yourself, “But isn’t the main problem that people invest emotionally?  Shouldn’t a mindful investor be able to time the market?”  The answer is: probably not.  Aside from any emotions you may have about the market on any given day, the basic problem with market timing is that you are trying to predict the future.  Look for example at the recent stock market as summarized by this chart from J.P. Morgan.

At each pull back in the ongoing relentless bull market, there seemed like good reasons why the market would finally start failing.  And each time, that “good reason” turned out to be an inconsequential subplot to the main story line propelling the stock market upward.  Remember the flash crash, the B.P. oil spill, the taper tantrum, the Ebola outbreak, and the oil recession?  When they occurred, they all seemed like excellent reasons for all hell to break loose, but it never did.  When you are immersed in daily events, it’s difficult to tell which events are truly monumental and which ones will simply be forgotten a few days later.

While in previous articles I emphasized the emotional aspects of trying to actively time the market during such events, there is also clear math working against even a robot trying to time the market.  The bloggers at Cash Cow Couple recently published an outstanding review of the hard math behind attempting to time the market over the short-term.  Here are a few key tidbits from that article (citations are provided there):

  • Several studies, including one by Nobel Laureate William Sharpe, determined the percentage of time a market timer would have to be right to beat a relevant index is something in the range of 70 to 85% of the time.
  • These statistics have nothing to do with the level of emotional intelligence of the investor and are driven entirely by math.  If you attempt to market time you must be in the market at the right time and out of the market at the right time.  If you fail in either task, you fail in general.  And a failure on a single day can put you behind the index return, at which point you have the even more difficult task of outperforming the index consistently to catch up.
  • Multiple studies have shown an overwhelming consistent failure to time the market by individual investors, media gurus, domestic and foreign investing professionals, and investing newsletters.

And as noted above, substantial market gains happen in just a few days during short, volatile periods within larger multi-year time spans.  The Cash Cow Couple cite a study by Professor H. Nejat Seyhun of University of Michigan that presents this graph.

 

If you could somehow miss the few worst months in market history, it would be a huge boon to your average annual returns.  More importantly, if you happen to miss the few best months, it radically cuts those same returns.  Choosing to stay in the market during these few best time periods while somehow magically missing the few worst periods is a near impossible exercise, even when making decisions entirely free of emotions.  It’s like standing in front of a shot-gun blast and trying to pick from mid-air a few favorable pellets, while dodging the malicious pellets.  And as discussed above, the actual performance data from real investors shows that almost no one performs this near impossible trick.

Long-term timing

As I discussed in the mindful bucket plan for “old” investors in Article 8.4, one of the best ways to guard your portfolio early in the withdrawal phase is to have a bucket of cash handy to invest after market crashes.  I also noted that we need to define the conditions under which it makes sense to deploy this cash reserve.  Technically, this is just a type of “market timing”, but it’s not the kind that people are usually talking about when they use the phrase.

The differences with long-term timing – The clearest difference between short and long-term timing is, obviously, the time scale.  The less obvious and more important difference with long-term timing is the limited context we are operating within for the mindful bucket plan for “old” investors.  Previously, I defined timing as buying AND selling stocks at the right times.

Crucially, the mindful plan for “old” investors only involves a one-time buy event using a predefined amount of cash, usually 20% or less of your portfolio.  Prior to that event, the mindful approach involves buying and holding starting in your younger years and continuing all the way to the start of retirement, no matter what the market does.  Once in retirement you only sell relatively small increments (something like 4% or less) on a predetermined timing, such as once a year, to fund annual regular expenses.  This predetermined sell timing and the associated amounts are entirely divorced from market gyrations.  You don’t care whether the market is up or down; you need money on a regular basis to buy groceries and live.  There is only one event (a large market crash) that would trigger additional buying of stocks under this plan, and only if the crash occurs in the first eight years after retirement starts.  No events trigger the selling of stocks based on market movements at all.

Unlike short-term timing, with this a very limited type of long-term timing, you don’t have to decide every day whether to be in the market, out of the market, or partially invested.  You only have one very limited decision point to consider.  You don’t have to be right all the time, you only have to be right once and only once.  Some readers may find this scenario so specific as to no longer fit within the definition of “market timing” at all.  That’s fine.  All I’m trying to do is make sure that this one buy event in the mindful bucket plan for the “old” investor is reasonable and feasible.

I would also argue that the longer time scale with long-term timing helps simplify decisions as well, regardless of the other limitations imposed by the overall investing plan.  With long-term timing, we aren’t attempting to make correct decisions every day, week, or month based on small-scale market movements.  Instead, we only need to make correct decisions about very large-scale market movements that may take many years to fully develop and may only happen once in a life time.

Switching Horses – In Article 8.4, I called this one long-term buy decision “switching horses”, because we switch from holding about 20% cash (or slowly spending that cash reserve down) to investing that cash reserve in the stock market along with the rest of our portfolio of stocks.  Aside from the more reasonable nature of long-term timing discussed above, we still haven’t determined when to switch horses.  Would it be after a 10% stock decline, 20%, 30%, 50%, or something else?  What’s a defensible number to trigger switching horses?

First, recall from previous articles (Article 4.2 and 6.1) that the path to positive stock returns almost always involves some dramatic temporary declines along the way.  We’ve seen that sometimes the stock market recovers quite quickly, and less often, it may take many years to recover.  Here’s a fresh example from the last 20 years, courtesy of J.P. Morgan again.

The price of the S&P 500 declined by about 50% from March 2000 to October 2002, but then recovered again by October 2007, about 5 years later.  The next price decline was nearly 60%, and the index returned to its former peak value again in about 5 years.

We can see there are at least two ways measure a stock decline.  The first is the duration of the decline and subsequent recovery.  The second is the amount of the decline.  The first metric isn’t much use as a cue to switch horses, because you never know in advance the total duration of the decline, and certainly not the duration of the recovery.  The second metric is useful, because we can wake up on any given day and calculate the amount of decline since the last market peak.  We won’t know how much further the market may decline in future days, but we can compare the decline so far to past data to determine when the decline has reached a level that is historically significant.

Here is one way to look at the history of stock declines from J.P. Morgan again.

The largest bear markets of the past have ranged between about -28% to -86% declines from prior peak prices.  The most recent nearly 60% decline during the “Great Recession” in 2007 was the third largest such decline in U.S. stock history.  These declines usually (but not always) take 1 to 3 years before a “bottom forms” and the subsequent recovery starts.  Here’s another way to look at it, again from J.P. Morgan.

This graph shows that intra-year declines are quite common and can often lead to a 30 or 35% price loss in just one year.  It’s interesting that in some years such steep declines don’t even necessarily result in a negative return for that same year, as seen in the 1987 example.  Looking across the last two graphs of long and short-term stock price history, we can see that declines of greater than 35% are significant in both time frames.  That is, declines of greater than 35% are relatively rare and generally signal the end of a major bear market is near.

I admit that this 35% threshold is not a statistical analysis, and it involves simply eyeballing the historical data.  Consequently, if you want to do a more rigorous analysis yourself, I wouldn’t blame you.  For me, this eyeballing gives me sufficient confidence to “switch horses” in my own investing plan when a 35% price decline occurs in the S&P 500.  My thinking goes something like this:

  • On a short-term basis, a 35% threshold avoids investing the cash reserve during most of the brief declines and quick recoveries in the last 35 years.  It avoids “false alarms”, where I might deploy my cash too soon.
  • For example, this threshold would have avoided investing the cash in 1987 (just barely), when there was a very quick recovery after the decline.
  • It also would have avoided investing the cash in all declines since 1980 except the big downturns in 2002 and 2007.
  • In the 2002 case, the 35% threshold was not triggered until the third year of the decline.  It would have avoided the scenario of buying well before the bottom was in, although it would not have picked the absolute bottom.
  • In the 2007 case the decline happened mostly in one year.  Again the 35% threshold would have resulted in buying stocks at a severe discount, although not at the final fire sale that was yet to come later in the year.
  • This threshold would have also avoided false alarms of relatively brief dips in:
    • 1961, when the recovery started in half a year
    • 1980, when the recovery started after two years

However, the comparison to 1946 is probably the toughest one.  As I presented in Article 8.3, after 3 years of declines leading up to 1946, it took another 4 years for the market to recover back to its previous peak.  So, in this case there were 7 long years where my portfolio value would have been going nowhere while my annual withdrawals would simultaneously have further depleted my portfolio value.  Nonetheless, the advantage of the mindful bucket plan is that the unused cash would be the primary source of the regular spending during most of such a 7-year drought.  So, the bucket plan would still provide a substantial cushion to the whole event; it would certainly have performed much better than having to sell stocks to fund spending for seven years.

No threshold is going to be perfect for every future stock market scenario.  But using the 35% threshold to deploy the cash reserves would provide a significant boost to the recovery of a portfolio in most of the past stock market declines.

Conclusions

We’ve determined that attempting to buy and sell stocks on a short-term basis (days, weeks, months) is futile and will almost always under perform a simple buy and hold strategy.  Long-term timing (over the course of many years) is fundamentally easier because it’s focused on very large historic events.  More importantly, if we constrain long-term timing to one critical decision (when to buy stocks with cash reserves), we can avoid most of the mathematical obstacles inherent to more general forms of market timing.

For the young investor, as presented in Article 8.1, the most mindful investing plan is to simply buy low-cost stock funds at regular intervals when long-term money becomes available, hold those investments until retirement (or similar spending phase), and ignore market gyrations entirely.  For the “old” investor who is near or in retirement, the mindful bucket plan, as defined in Article 8.4, is a feasible and prudent approach to minimize “sequence of return risk” and protect your nest egg.  That plan involves reserving a relatively small bucket of cash (about 20% of the overall portfolio) to invest if/when a large market decline occurs, which we have now defined as a decline greater than 35%.

From an emotional perspective, investing cash after a 35% market decline will be a difficult decision, particularly when you will have no idea how much more the market may plummet.  However, such an approach is exactly in line with what we learned in Article 4.1.  The best times to invest are likely to be those situations that make you the most queasy and fearful, consistent with the “be greedy when others are fearful” maxim from Warren Buffett.

Rigid use of the 35% threshold is also going to help execute the plan.  Sticking to a plan where you don’t invest at a 34% decline but go ahead and invest at the 35% mark is somewhat arbitrary.  However, it provides a clear rule to follow, which also helps keep emotions out of the decision process.  As we have discussed before, having a clear plan is one of the most important tenets in investing and much better than a plan where such rules are vague or undefined.