Investors who collect reliable information and time their actions to predictable shifts in asset-class return cycles on average will outperform investors who buy and hold for the long run. Simply, they will make hay while the sun shines, just as above-average farmers make the most of good weather and clear skies.

This is a nearly one-to-one parallel, as I see it: How the farmer goes about the process of gathering information, building a forecast, and acting on that forecast mirrors how cyclical asset allocation is properly performed.^[1]

At the outset, one would expect the above-average farmer to be well acquainted with his crops, land, pests, insecticides, seeds, and machinery. He also knows what the quality and quantity of his product should be if he is to earn the most when it is shipped to market. All this puts him in the game, so to speak. However, if he consistently is to be an above-average performer, meaning that he will more often than not produce a high and quality yield, he arguably will time his actions in the field with admirable precision. Primarily, he will excel at coordinating his actions with the weather.

Of course, the good farmer working in the field keeps his eyes open. He watches the skies and the clouds, and he knows that storm clouds, in particular, are packed with information. When clouds gather on the horizon, he estimates how long he can continue to plow before the rain will stall his work. By checking the clouds against the wind and appraising their shade and shape, he attempts to determine whether he has hours or just minutes to get his equipment in the barn. This is a judgment that experience brings; the farmer’s practical experience allows him to make a forecast and act on it.

But in most cases, the information he collects with his eyes and the application of that information are not enough to ensure an above-average yield year to year. He must rely on a bit more information than that.

Weather shifts predictably, although not perfectly, with the seasons, and it shifts less predictably week to week and day to day. This requires the good farmer to be a forecast collector for both the longer and shorter terms. Seasonally, and perhaps weekly, he checks in with his local agriculture forecasting service, the people who track the weather diligently and scientifically, and every morning he listens to the weather forecast on the radio or watches one on TV. Each of these forecasts helps him determine the most favorable times to plant, pick, water, fertilize, and spray.

At this point, the information carried in a forecast, the reliability of that information, and personal experience all converge to bring about a conviction to act. When that conviction level rises, timing becomes more precise.

Specifically, the farmer adjusts his actions in the field based on a sum-total rendering of the forecasts he collects—whether they are derived from advanced scientific evidence, the analysis of a local weatherperson, or his own eyes and experience. Additionally, none of these forecasts is a sure thing; at times, each will be off. In theory, if a weather forecast is always right, the farmer always will know what to do. He similarly always will know how to act if a forecast dependably is wrong. But the more real-world example is one in which a forecast is right only some of the time.

Let’s say, on average, that a particular weather forecast is right only 67 percent of the time. In this case, the farmer has to make a judgment call on whether to act on this report. He also will want to know if there are additional pieces of information that will help him improve that forecast.

The timing of the forecasts themselves also matters a great deal.

Let’s say a forecast is for sustained stormy weather. If received just before planting, the forecast might cause a delay in planting; if received during the harvest, it might cause an acceleration of the harvest; if received in between planting and the harvest, it might elicit no response whatsoever. Thus, the extent to which a forecast causes a change in behavior depends on both the historical accuracy of the forecast and the stage of the crop when the forecast arrives.

These are the variables the good farmer must weigh as he builds a reliable plan of action. And these are your variables, too:

Farmers must serve Master Weather just as good investors must serve Master Shock. Yet each of these masters can be managed. The farmer develops a weather forecast by collecting information from a range of sources, and he times his actions in the field based on his experience with that information. He must know his crops and equipment to get in the game, but he must know them in relation to the weather if he is to apply them with the best possible effect.

Similarly, good investors must know the characteristics of the separate asset classes. If they are to allocate to them with the best possible effect, they must understand them in relation to the range of powerful macroeconomic variables that include taxation, monetary policy, interest rates, and inflation.

Timing, in all its investing shapes and forms, is a dirty word to those who have little faith in the various forecasting methods that attempt to predict short-term swings in asset prices. But here I ask a basic question: Why do asset classes swing away from their long-run trends for shorter periods of time?

The most obvious answer is that enough investors from time to time act in the same way and thereby cause pronounced swings in the price of distinct groups of assets. In other words, enough investors will have bought or sold enough of the same types of investments to cause a significant price movement in those investments.

Now I ask a follow-up question: Why do enough investors suddenly act in the same way? Is this randomness? Coincidence? The result of a cause that even if known today was not predictable yesterday?

I’m not so sure it’s any of these. In fact, when enough people act in the same way at the same time, I am from the school that believes there is not only a tangible or identifiable reason for this collective action, but that the reason might be behavioral and predictable.

A simple marketplace example can describe this.

Surely you are familiar with retail come-ons, such as “Shop Our Blowout Memorial Day Sale: All Prices Slashed!” or “Only Ten Days Left to Shop Our Incredibly Low Prices!” Stores preannounce sales, or even price increases, all the time. But the behavioral effect of the former, the preannounced sale, is much different than that of the latter, the preannounced price increase.

For starters, few of us will shop a store the day before a preannounced sale—the only ones who might are those who are ignorant of the coming sale or for some reason must have an item in that store on that particular day. And who are these full-price shoppers? Well, whether they are ignorant of a pending sale or not, if they must purchase an item at the going price, despite the fact that it will very shortly lower in price, we can say they are inelastic consumers. Meanwhile, those who possess full knowledge of an oncoming sale and have flexible schedules will be able to take advantage of lower prices when the sale goes live.

As for the preannounced price increase, rather than a timed delay of activity, we will see a timed acceleration. If prices in a store are to increase tomorrow, consumers will have a strong incentive to purchase items in that store today. As expected, only those with inflexible plans and/or a lack of knowledge of a price hike will shop a store on the day of that hike.

Consumers—at least, the elastic ones, who have knowledge of a pending price increase—will bask in the sunshine of the current lower prices; they will purchase more before the price goes up. Similarly, consumers with knowledge of an upcoming price decrease will wait for the sun to shine; they will plan their purchases for when prices go down.

When it comes to prices—and saving a buck or making a buck—strong behavioral forces determine group action. Indeed, timing happens all the time, and it is often a predictable event.

This analysis easily graduates from the consumer experience to a macroeconomic shock that preys on human behavior, the direction of the economy, and the price movements of major asset classes: tax policy. Although tax increases and tax decreases are different animals, they are alike, in that they usually are preannounced. Hence, each event can be the basis of a forecast that is very accurate.

For example, when new tax legislation passes, anyone can know the exact date that the tax change goes into effect, with that date either representing the beginning of a storm or a clearing of the skies. There’s a nuance to this, however. In the case of tax cuts, legislation can be phased in; in the case of tax increases, the current lower rates can be phased out. But the actions of taxpayers in all cases are predictable: As a group, they will avoid the higher-tax storms and bask in the lower-tax sunshine.

A very clean example of this occurred during the second term of Ronald Reagan’s presidency. Back in 1986, on the administration’s prompting, Congress lowered the top marginal income-tax rate to 28 percent from 50 percent, with this cut scheduled to be phased in over a two-year tax period. At the same time, the new legislation increased the capital gains tax rate to 28 percent from 20 percent. So we had a preannounced tax increase and a preannounced tax decrease, indicating two different periods when the sun presumably would shine on taxpayers.

In the case of the preannounced tax increase, investors with accumulated capital gains largely chose to realize them before the capital gains rate was lifted to 28 percent. This is an obvious case of making hay while the sun shines, although there is more to the story.

Because the prospect of higher capital gains tax rates in the future led to more people paying taxes at the current lower rate, more taxes were paid, which fattened the tax-revenue coffers at the federal and state levels. In some states, officials projected these revenue gains to continue, and in doing so, they mistook a one-time windfall for a permanent event.

In my own California, for instance, spending was increased to match the higher revenue projections. And when the higher revenue did not materialize, the state went into deficit. To make matters worse, instead of cutting back on spending, California raised taxes and pushed the state into recession, all while the federal economy was also slowing down. (Thinking back to the location effect, it’s not surprising that California as a whole and its small-cap companies, in particular, suffered a worse fate than the average of the other 49 states.)

Tax changes have far-reaching and predictable implications, and we can view this again in the case of the Reagan income-tax cuts of 1986.

These cuts were phased in over two years, a period that coincided with an economic slowdown. And why did the economy slow? Our store example can help answer this.

Recall that when stores preannounce sales, consumers will delay their purchases until the sale prices take effect. This delay will cause store sales to slump. And taxpayers react to incentives or disincentives just like consumers: They will delay activity if they know they will profit by doing so. Because the Reagan income-tax cuts of 1986 were phased in, there was a negative impact on behavior from the consumer level to the corporate level in the short run. In economist terms, the postponement of economic activity at the consumer level and the juggling of the books on the corporate level had the effect of delaying income recognitions—but only until the full rate cuts went into action.

And once they did, guess what? The economy boomed, with the GDP gaining by double digits.

If consumers and taxpayers can time their behavior with precision to periods when the returns will be greater, to the periods when the sun shines, there’s no reason investors can’t also. The challenge then becomes to detect, or forecast, that sunshine. We begin this process by searching for patterns of relative performance between stocks and bonds.

I have stated that the stock/bond (or equity/fixed-income) decision within a portfolio is the most important one an investor can make. The reason is simple: Bonds produce lower returns with lower risk, while stocks produce higher returns with higher risk. Everyone wants higher returns, but the level of risk we can stomach varies greatly person to person. Hence, so does the level of returns.

However, if we can discover persistent patterns of relative performance between these bellwether asset classes, it stands to reason that we can mitigate much of the heartache associated with this choice while also maximizing our earnings over the long haul.

So let’s take a look.

Pillaging through the quarterly returns of the S&P 500 (my stand-in for the performance of stocks) and the ten-year Treasury note (my stand-in for the performance of bonds) for the last three decades, I identified seven full cycles in which bond returns systematically outperformed stock returns (see Table 12.1).^[2] These “fixed-income” cycles are about equally divided between those that lasted 6 quarters and those that lasted 12.

As an aside, perhaps the distinct duration of the cycles lends support to the idea that election cycles at the congressional level (every two years) and the national level (every four) affect the direction of economic policy and, thus, cycles of relative performance in the stock and bond markets. This tidbit adds some grist to our cyclical mill. But more to the point, the fact that these relative-performance cycles persist at all means there is a possibility of developing a strategy to exploit them.

Let’s first look to a usual suspect: recessions.

It is sensible to expect stocks to underperform bonds during recessions, when the GDP is contracting, and outperform during recoveries, when the GDP is expanding. Simply, investors will protect themselves with safer bond or cash positions—or fixed-income positions—when the economy turns sour.

Yet the recession record as it relates to fixed-income cycles is not perfect. According to the National Bureau of Economic Research, in only three of our fixed-income cases was the business cycle either all or part recessionary, while in four cases, the fixed-income cycles coincided with expansions. Thus, the data shows conclusively that economic recessions are neither a necessary nor sufficient condition for fixed-income cycles.

Is inflation our indicator? Again, the correlation is not perfect: Fixed-income cycles have occurred when inflation has trended both upward and downward.^[3] However, when you subtract the inflation rate from the Treasury bill yield, essentially removing the effect of inflation, you get what is known as the real rate. And in our study, each time a fixed-income cycle occurred, the real rate was falling.^[4]

From this result, we can draw a quick investment rule:

The phrasing of this rule is important. Even though a falling real rate is a good indicator of fixed-income cycles, I cannot conclude that a falling real rate is bullish for bonds or necessarily bearish for stocks. That’s because there are different degrees to which stocks and bonds, as distinct groups, perform cycle to cycle. Indeed, when the real rate is falling—a generally nice time to be invested in fixed-income instruments—stocks also have performed well enough.

This brings us to the concept of valuation, what we estimate the relative value of assets or asset classes to be. When forecasting how stocks and bonds will perform in the near term, we want to grade the economic conditions from least to most favorable, a process that brings economic policy and economic performance neatly together.

Likely as many economic forecasters exist as weather forecasters. And they each select from a broad basket of data when developing their forecasts. For the economic forecaster, these data points might include consumer prices, interest rates, corporate profits, commodity and producer prices, currency exchange rates, unemployment rates, productivity and wage levels, and on and on. They also look at exogenous factors that are outside the control of individual investors and the collective financial markets. These might include war and peace, national strife and passivity, or the effects of unexpected disasters, both manmade and natural. However, despite the breadth of data at their disposal, many economic forecasters put the greatest weight on just two of their conclusions: the future direction of inflation and the real (or inflation-adjusted) rate of growth for GDP.

Table 12.2 describes the level of stock and bond performance in relation to the direction (increasing or decreasing) of each of these variables, inflation and GDP. The results are based on monthly data from 1948 through 2004, and they make a lot of sense.^[5]

Stocks performed best when inflation was on the decline and the GDP was increasing, what we might call a bullish economic environment. Turning the tables, when inflation was on the rise and the GDP was declining, stocks fared the worst, although, as a group, they did turn in a gain of 0.22 percent. Bonds did the best when inflation and the GDP were both decreasing, and the worst when both were increasing.

Fifty-six years of data is a nice sample, and from it we can draw some pretty firm conclusions about how stocks and bonds will perform in relation to a very simple rendering of the economic environment:

Stock and Bond Performance in Relation to GDP and Inflation Growth

It stands to reason that if you are to take advantage of these rules of thumb, you want to possess a reliable forecast of the directions of inflation and GDP. However, such forecasts can range in their accuracy from excellent to quite poor. A good explanation for this disparity is that often economic forecasts do not account for public policy.

In particular, shifts in tax rates and the procedures the Federal Reserve use are greatly related to the future direction of GDP and inflation. A forecast that does not account for these shifts stands an excellent chance of being wrong.

For instance, lower tax rates on workers will increase the incentive to work more, earn more, and put those earnings into action, while lower taxes on business and capital will bring on business expansion and higher employment. This is economy-stimulating stuff that results time and again in a rising GDP. What good is a forecast that misses the impact of such policy?

Meanwhile, a Federal Reserve that adjusts the amount of money in circulation by reacting to the direction of prices in the overall market stands the best chance of holding inflation in check. Here, the thinking is that whenever prices rise above a certain target range, too much money exists in circulation, and when prices fall below that range, there is too little money in the pipeline. However, the Fed can control this simply by tracking prices in the aggregate and acting appropriately on that information.^[6] Nevertheless, a sound inflation forecast must attempt to discover the direction of Fed policy, regardless of the sensibility of that policy.

Because you can read the newspapers and watch TV, you can monitor fiscal and monetary policy on your very own. And because these policies are excellent harbingers of the GDP and inflation, respectively, you stand a good chance of formulating reliable economic forecasts. But there’s nothing like a second opinion, particularly an expert second opinion.

Just as seasoned meteorologists develop forecasts that are correct more often than not, the more experienced top-down economic forecasters boast good track records when it comes to anticipating behavioral changes in the market. Even better, major newspapers regularly poll a variety of these analysts, putting together what they call “consensus” forecasts on the future direction of the stock market and economy.

How good are these forecasts?

As a financial advisor, I formulate my own forecasts, and I also sit on financial advisement boards that construct consensus opinions on the economy quarter to quarter. These forecasts have missed at times, but they have been right much more often than not. This is good news, not just for me, but for every investor who is grounded in the forces of supply and demand.

I make this last point because we can profit from an economic forecast only if we can apply it to a proven investment framework. And the most provable one I know is that which projects public policy and the range of macroeconomic variables forward to the supply-and-demand responses they will illicit. This is the framework I set forth in this book. If you subscribe to it, it is the framework to which you will apply the economic forecasts you both build and collect.

Here’s an example of how the procedure might work:

At the start of 2005, the Wall Street Journal’s consensus economic forecast called for a 1.4 percent increase in the short-term interest rate (Treasury-bill yields) and a 0.89 percent increase in the long-term rate (ten-year Treasury-bond yields). Bond prices are inversely related to yields, so the predicted increase in yields was a bearish signal for bonds. The Journal also called for a low 2.5 percent inflation rate and a strong 3.6 percent GDP growth rate, an environment that historically has favored stocks over bonds.

The Journal’s consensus was bullish, to say the least, although this forecast historically has not always been correct. But if in 2005 you understood the relationship between economic policy and the relative performance of the separate asset classes, particularly the stock/bond split, you would have been just as bullish.

In the six quarters prior to the start of 2005, the GDP had been trending upward, a response to the 2003 tax-cut package that included lower rates on capital gains and investor dividends. Few, if any, signs in early 2005 indicated that this period of sunshine was about to end. Other data trends included lower unemployment, rising corporate profits, strong productivity, and record high tax revenues—all marks of sustained economic expansion. Finally, on the monetary front, fears were bubbling that higher future inflation was a threat. Yet, regardless of whether it was, all signs pointed to the Fed holding it in check: The Fed had begun the anti-inflationary procedure of raising interest rates six months earlier, and most statements from the Fed said this process would continue in a measured way.

This confluence of indicators should have had you investing properly. If you read the papers, understood the relationship between public policy and future economic realities (in other words, if your framework was correct), and backed up your opinions with a well-known consensus forecast, you might have favored stocks over bonds during 2005. And you would have been right to do so. For the year, stocks as a group gained 4.2 percent, while bonds fell 1 percent.

We live in a world of a tax hike here and a tax cut there, of a year when the Fed and other central banks managed inflation handily and a year when they made a mess of the monetary works. As a result, we live in a world of bond moments and stock moments with many shades of stock/bond gray in between.

But in most countries, the legislative process is long and drawn out, so there is plenty of time for anyone so inclined to figure out what policies are coming down the pike. And although it is true that final legislation is never identical to initial proposals, policymaking is rarely an impromptu activity; in most cases, final policy law is related to earlier bills that have long been discussed.

In this way, “timing,” as it relates to cyclical asset allocation, is not an onerous process—it is not day to day. This is a good reason why big investment firms make a habit of meeting quarterly to decide their major allocation moves: They understand that the markets shift in relation to significant world events. And although they make daily allocation decisions based on the minutiae of market data, for the most part, they react to the “big stuff” at quarterly intervals.

Anticipation is involved in this process, and anticipating cycles is a better strategy than reacting to cycles: When you anticipate, the window in which you stand to make above-average gains stays open that much longer. This strategy, however, is not without peril: Sometimes your forecast will be wrong and you will have to adjust. When unexpected macroeconomic shocks or unanticipated policy changes occur, ones that have you suddenly poorly allocated within your portfolio, you should cut your losses, correct your forecast, and allocate to the new cycle.

That said, more often than not, your forecast will be correct if you perform the due diligence described in this chapter. Doing so also will have you improving your performance year to year and cycle to cycle. Overall, if you behave like the above-average farmer who merges appropriate action with forecasted swings in the weather, you both will make hay while the sun shines and protect your portfolio when storm clouds threaten.