Tag Archives: investing

Time for giving thanks

Leave a reply

(No forecast, just words.)

I am especially thankful that none of my forecasting models can read the newspaper, or even know what I think.

The only things the models know are long established reputable economic data series. Each data set has been publicly available for at least 10 years; most have at least 3 decades of experience. A few go back much further. Each variable has a proven statistical relationship to at least some stock market prices. (Mimimum 95% confidence level.) Nearly two decades ago the models started with about a dozen key economic variables, but now roughly 4 dozen macroeconomic variables are included along with millions of data elements from individual stocks. Variables still are almost entirely focused on the US stock market. Things started with a single macro model that went unchanged for years. Now there is an AI swarm of thousands of models, each with a slightly different point of view, and each having to prove its current value and accuracy profile. Each model gets refined on a weekly basis. Every forecast for any stock results from a minimum of 30 independent analyses.

I started these stock forecasting models because I was absolutely horrible at picking and timing stock purchases. My heart, emotions and hunches always got the best of me. Partly for that reason, there are no parts of the models that include my own expectations (like on interest rates or corporate profits in the coming month). Very pointedly, however, I do include a number of variables that do have a high ’emotional’ quotient. The VIX market ‘fear gage’ is one example. Numerous long-standing surveys make the cut as well.

My point in saying all of this is that over the next year or so it seems highly likely that individual stocks and the stock market will have an unusually high level of uncertainty. The stock market DOES read the newspaper and dwells on rumors and false trends. I expect to see unusually high volatility. Should that occur, the models reported on here will almost certainly predict a reversion back toward trend.

In the end, I trust the stock market will usually do what it usually does in response to changing economic circumstances. That’s what it has done for the past 17 years, at least. Who knows, maybe all of that will change, and pigs can start to fly. Perhaps, but I have no evidence of that.

We are in the early stage of what promises to be a major US stock market bubble. This is unlikely to be especially high and sustained real economic growth, just pure bubble. And it could last for a couple of years.

Wall of Worry. Bubble Starting?

2 Replies

Forecast October 2024 thru March 2025: Climbing the Wall of Worry continues. My oldest and most accurate long term models are skeptical about the market, and have been negative for months. These econometric models can never forget the plunging prices of past stock market crashes. That’s their job. They are supposed to worry.

This time might be ‘slightly’ different. I won’t be surprised if the models are overly cautious for several more months, and the market keeps gaining.

The U.S. stock market is climbing a “Wall of Worry“, a classic stock market expression that applies at about this time during every business cycle. Stock valuations have been rising for several years and now are high by nearly all measures. But, prices keep rising anyway. ‘Everyone’ knows the market is overdue for a tumble, but ‘nobody’ wants to miss out on the relentless gains that stocks keep washing in. Eventually ‘all’ markets do falter, and most speculators are completely surprised. My most recent and detailed models say the end may not be imminent, At least, not for the next few months.

At the request of one of my few readers, I am bringing back my oldest and ugliest chart to illustrate. My wife calls it ‘garish’ and she is never wrong.

The graph below displays all of my actual 6-month market forecasts going back to 2007. The shape of the graph is the S&P 500 plotted on a logarithmic scale. Each dot is color-coded with the dot color indicating the approximate forecast value for the coming 6 months. (So, ideally a big market drop should follow a black diamond, and a strong market rise should follow a big green triangle.)

The message here is that the negative (red) forecasts of the past several months paint the wall of worry, just like they did through the pandemic mini-bubble. Eventually, they will probably prove to be approximately correct. At some point not too far away the market will probably be cheaper than is is today.

The plot (below) of a long-term model of the S&P 500 shows what certainly looks like the start of a new market bubble. Since at least 1871, the growth of U.S. GDP and the U.S. stock market have corresponded quite closely. That is no surprise: GDP tries to estimate the total volume of goods and services, while the S&P corresponds to a big chunk of that same economic output. The trend line here plots the current (and future) path of the S&P 500 based on the ongoing relationship between Real Potential GDP and long term interest rates. It is far from a perfect model, but it shows a growing divergence: a new bubble? (This model shows the S&P 10% above trend, another model says 14% above trend.)

Here is a somewhat more positive look at a series of market index forecasts. The forecasts span from 1-week through 6-months. The models that produce these plots are based on gobs more data than my original models and are not quite as ‘fixated’ on the market crashes of the past. For good or ill, they are more trusting of current economic data.

If the Orange Octogenarian does something exceptionally bizarre in the next month, then all bets are off.