Do You Have a Market Evacuation Strategy?

Here's a piece from US News that says we all need a market evacuation strategy. What exactly is a market evacuation strategy? It's when you sell your investments and go into cash -- just before the market takes a big dip. Then once it's dropped, you get back in and buy at the bottom to make big bucks.

Color me skeptical. Isn't this what people have been trying to do (and failing) since there were markets? That said, let's hear them out.

Before we get to the "how's" of this "strategy", let's cover the "why's". Here's what the post says:

Examining stock market history, you can see how the market goes straight up for a particular period of time (five to seven years) and then suddenly, your portfolio seems to drop like a hot potato. This has happened so systematically that it makes the market seem like a well-oiled, broken machine.

And when they say "drop", they mean it -- noting that the market will have losses of 40% to 60% every five to seven years.

So, how do they suggest determining when the market will take a big fall? They say that "using technical information about the market, you can draw some conclusions about when the market is so overpriced and out of control that it is about to drop precipitously." Here are the four technical analysis tools they say you should keep your eye on:

Momentum. What is happening with market prices in comparison to the 200-day moving average or the 10- or 12-month moving average?

Yield spreads. As the market begins to show signs of weakening, fixed-income traders will begin to pull out of riskier bonds, such as high yield, and go into safer bets, such as government bonds. At that point you will begin to see a larger disparity between the rates on those two kinds of assets.

Volatility. The way to measure volatility is to keep an eye on that option pricing by looking at the VIX, the Volatility Index, also known as the "fear gauge."

Correlation. When riskier assets get sold en masse, the correlation of those various assets begins to move closer together. Assets that would historically move in very different directions begin to move together in almost lock step.

They conclude by saying that all four of these need to be pointing in the negative direction at the same time before you should make a move.

Here's what I say:

• This is a GREAT idea! You beat the drops and yet capture the upsides. If you did this for only a couple cycles and with any reasonable amount of money, you'd be fabulously well off.


• The only problem -- no one can do this consistently and accurately. If people could, they would (of course -- I would if I could.) But they can't.


• Can you at least get close? Maybe, maybe not. I know some commenters here will say that they can (and have) get out before a drop and then get back in before the big upswing, but my experience is that the vast majority of people will get out too early/late and get back in too early/late, hurting their performance way more than they'll help it.

What's your take on this issue? I'm sure there will be advocates on each side of the topic, so I'm interested to see the discussion.

Comments

[DIY Investor]

I defer to Benjamin Graham on this one

If I have noticed anything over these 60 years on Wall Street, it is that people do not succeed in forecasting what’s going to happen to the stock market.
-Benjamin Graham, coauthor of Security Analysis

[Ganesh]

For a long time, I have sat in the 'keep the asset allocation constant' camp. By this I mean I look at my asset allocation periodically and adjust it based on a set equity percentage (80% for me).

Lately however, I have been toying with the idea of modifying this equity percentage on the basis of some market metric. I chose the easy way out, looked at P/E which has very good 'reversion to the mean' characteristics. Only time will tell whether this is a better approach but it does tell me to buy when the market is tanking and sell when the market goes up.

What I do - I basically determine my equity percentage within a 65% to 95% band based on the percentile of the P/E ratio based on a 5 year period. I usually look at a fortnightly or a monthly re balancing.

[Alex]

My take is this will not work. There is no way to predict what will happen. The market behaves in unexpected ways. In addition, we've had events over the years (i.e. 9/11) that are completely unexpected that can tank the market. It's the very act of people trying to get out that makes the market tank in an event like that. I believe most people are better off picking a solid allocation and letting it ride.

[tllstaco]

We used to follow this strategy and we could never reliably predict the down turn or the up turn. Neither could the multitude of brokers/financial advisors we trusted. A few years back, we decided to switch our portfolio to more of an income generating portfolio. As a result, we don't necessarily want to sell into the downturn since the reinvested dividends/income buy more shares during that time. More shares mean more income. We try to buy more equities when we think the market is down and most people are pessimistic.

[tom]

In theory, great idea.

In reality, terrible idea.

Markets are completely irrational and driven by emotion. You can no more predict someone's reaction to an event than the rapture itself. They only thing you can predict is that, en masse, people will react irrationally.

Here's the rub, investors don't need this! Why the heck do you need an evacuation strategy when you have an appropriate asset allocation? If you have an appropriate AA for your age and risk tolerance, you'll never need to pull out of the market; you'll easily absorb the bears and make bank on the bulls.

If you do not have an appropriate asset allocation, then you're foolish. Why would you risk your hard earned money?

[Everyday Tips]

My take is that is a ludicrous plan. Sure, sounds great, and it is a lot easier to see trends that already happened on a chart. Looking at a chart and predicting the future is a lot harder.

If you can't handle the ups and downs of the market, then stick with safer investments.

[Mike Hunt]

I'm waiting to see Old Limey's response. He's been fortunate, skilled, and wise enough to do this twice successfully.

-Mike

[crashdamage1957]

I agree with the above comments. Asset allocation is a strategy; trying to time the market is a fallacy.

[optionsdude]

You can't time the market. It is as simple as that. You just have to have some sort of strategy that allows you to take some money off the table after a big run to have cash available when stocks turn again. I usually use put options or sell a few shares at a time on the way up and buy a little bit on the way down. It then works out that my overall selling price is greater than my overall purchase price.

[Apex]

40-60% every 5-7 years?

I didn't read the article but this would be pretty easy to succeed at if there were these kinds of corrections with this regularity. Unfortunately if they actually said that they are full of crap.

Lets assume you could pick the top and the bottom (no matter how good you were you would still miss these by several percentages).

Here is a chart showing market corrections since 1929:

http://www.ritholtz.com/blog/wp-content/uploads/2009/02/bear-markets-comparison-xlrg.gif

If we start right at the worst bear market in history, how many bear markets have lost 50 or 60%? One, 1929. How many have lost 40%? Three, 1937, 1973, 2007. How many have lost 30%? Three, 1939, 1968, 2000. There were 5 that lost 20% but at that point when you miss 5 points on the top and bottom, what have you really gained and that assumes you can get it right.

So over 80+ years there have been 4 corrections of 40% or more (only one of 60%) and another 3 of 30% or more. So a 40% correction is happening every 20 years, not every 5. A 30% correction is happening on average every 12 years and a 20% correction maybe every 7. But another thing is worth noting based on the years.

1929, 1937, 1939
1968, 1973
2000, 2007

See a pattern? The corrections happen in bunches. There is no regular pattern of any reasonable correction every 5-7 years. There is a decade of corrections, then a fairly large gap, then a decade of corrections, then another large gap, then a decade of corrections. Some of the 20% corrections also cluster around those decades and make them last more like 15 years instead of a decade.

So if this article had merit it would be entirely refuted by the fact that the corrections do not happen with any recognizable pattern of regularity. They cluster around bad times. We are currently going through one of the 3 bad times of the last 80 years. I don't know when it will end but when it does, history gives no reasonable guide that you can expect a 40%+ correction 5 years later.

The article is built on an entirely false premise and is thus bogus.

[FMF]

Apex --

Here is their exact quote:

"Tanking is not a loss of 10 or 20 percent, although no one is happy when they lose that amount. True tanking is a 40 to 60 percent loss–the type of losses the market has, and will continue to have, every five to seven years."

[Apex]

@FMF,

WOW, that is just blatant hyperbole by them. Tanking has not happened every 5-7 years. In fact it never happened once from 1980-1999. Most of their readers are probably old enough to remember that too. Just because it happened twice in the last decade is no reason to ignore the previous 2 decades. That is some pretty poor journalism on their part.

[mysticaltyger]

@ Apex....Actually, I think the market did tank in 1987. It didn't lose money for the year, but some time in October 1987, it lost 22% in one day, didn't it?

Overall though, I don't disagree with your argument.

[Apex]

@tiger,

Yes it did, but the article made it clear that their definition of tanking starts at 40%.

And that was a 1 day event, there was almost no warning, no rolling over, no chance to get out. So unless you had a crystal ball how was one supposed to predict that decline. And if one is to argue the market was over-valued, well it regained all of its loss 9 months later. It went from 2,200 to 1,700 and then from there it went back to 2,200 in 9 months and then straight to 11,000 over the next 13 years. So if it was over-valued at 2,200 then I guess anyone who would have gotten out on valuation would have stayed out for all of the remaining 13 years, unless of course 9 months later 2,200 was suddenly cheap somehow.

1987 is an interesting anomaly but it is simply a larger version of the flash crash we had in the last year. It is not a bear market tanking.

[Matt]

Nope.

I did not panic sell 2 1/2 years ago becasue it was oversold and it came back. My new thing is if I move money I move it slowly and thoughtfully without panic and keep a balance on things.

I am also keeping closer tabs on my risk exposure.

[Concojones]

Timing the market, been there, done that... never again.

In late 2008, I almost bought because from a valuation point of view, a number of shares were dirt cheap, screaming buys. Thinking I could time the market (after all, I had expected the 2008-2009 crash), I decided to wait another year for "the real bottom". Result: I lost my chance, and I drew my conclusions: don't buy based on market timing, only based on valuation. I have applied that and haven't looked back since (I have bought shares that seemed cheap despite my expectation of a new market correction).

@Mike: from what I remember Old Limey posted here, his 2008 exit from the markets was a coincidence.

[Old Limey]

Nothing that you read about what the market is going to do next can ever be relied upon. For one thing there is a whole lot more market manipulation happening these days by the biggest players that include the Federal Reserve Bank with their POMO injections into the market through the biggest investment banks such as Goldman Sachs and others. Prediction has never been more difficult because of the speed that information now whizzes around the globe 24/hours day and the way that countries are becoming more and more interconnected because of the big fluctuations in commodities and currencies and the worldwide debt problems that seem to monopolize the financial news.

I don't try to time the market, particularly on a short term basis. I just use "Fund Selection" to try to keep the 1/3 of my money that is in mutual funds always trending upwards. That's where having a database of 11,000+ mutual funds and powerful ranking software is essential. The other 2/3 is in CDs yielding an average of 4.93% tax deferred and in Muni Bonds yielding an average of 4.85% tax free that are all being held to maturity. To find the best bond funds for what I want, I first rank by a measure of volatility, and then starting with the least volatile I scan down the list looking for the few funds that really stand out as having very low volatility as well as really nice annual rates of returns. The funds that I find are invariably Institutional funds that have large minimum purchases, typicaly $100K, $250K, or $1M.

I believe that the reason the Institutional funds seem to perform better is that their assets under management are large and that they have many assets that they have owned for quite a few years that are producing higher yields that can be obtained by buying new issues. It's the same with the CDs and Muni Bonds that I own, you can't buy them at the prices I paid because interest rates have dropped so much in the last few years.

Of course, if I was a young man that still needed to strive to produce large capital gains in the stockmarket, what I am doing wouldn't be of any help since I am now very satisfied if I can compound my portfolio by 5%-6% per annum year after year while minimizing my income taxes. Nothing I own is traded on the stockmarket. At the end of 2007 the four stockmarket indexes that I was then using to appraise stockmarket health had all gone negative so I got out of the stockmarket and into CDs and Bonds back then, and haven't been back in since. The four indicators were the running summations of "New Highs - New Lows" and "Advancing Volume - Declining Volume" for the NYSE and the NASDAQ.

[V]

Timing the market is possible, but it's not easy.