Most Thankless Well-Paid Job in China: Market Forecaster

Calligraphy from China First Capital blog post


Looking for a new career with plenty of growth potential, and low standards for success? Here’s one to consider: China market forecasting. Rapid economic growth and urbanization are both creating huge demand for market research predicting future areas for opportunity and profit. Pay is good. But, there’s another aspect to the job that will appeal to many: repeated failure is no obstacle. 

Market research is, of course, a treacherous profession anywhere. Predicting the future always is. But, in China, market forecasting is particularly hard. It’s mainly been distinguished by how often, and by how much, the predictions turn out to be wrong.  Market segments in China grow so quickly, so explosively, that it makes a fool of just about anyone trying to guess its economic future. 

I’m reminded frequently of this these days. We’re working on a complicated infrastructure financing. One of the central components of the deal is a now two-year-old forecast of car purchases and driving patterns in China. The forecast was prepared by a respectable outfit in Hong Kong, and my guess is that they charged quite a lot to do it. But, looking at the numbers now, they seem ridiculous, like numbers pulled out of thin air – which is probably what they were. The actual growth of car traffic and car purchases over the last two years in China has been much higher than these predictions. In other words, the forecasts weren’t off by a mile, but by a light year. 

Given that track record, it’s surprising these market forecasters can continue to pay the rent, let alone prosper. And yet they do. It’s a familiar paradox: we know projections are often wrong, and yet many business decisions, often with billions of dollars at stake, are made on them. It’s probably connected to what’s sometimes called “the scientific theory of management”, which tried to systematize complex business decisions into quanta of data.

It’s the same approach taught in business schools, and is certainly one of the reasons so few MBAs make successful entrepreneurs. A hunch is often a better tool in business than a spreadsheet. Indeed, I’ve yet to meet a successful entrepreneur who ran his business, or started out in life, based on a market forecast. 

In our case, we’re stuck using the projections on auto traffic, because there’s nothing else available. So, we send them out to investors with the guidance to take the projections with a grain of salt. If not a fistful. This creates its own set of problems, including frequently the request to do a new set of “up-to-date” projections. In other words, the solution to bad projections is – you guessed it — to commission more projections. As I said, it’s a great job, being a market forecaster. 

The errors in a bad projection become cumulative. The longer the time line, the more distorted the projections will usually become.  In our case, we’re using a 25-year projection. So, these sizable errors in the first years will propagate across time. Year by year, the forecast becomes less and less tethered to reality, like the NASA space probe that escaped its flight path, lost contact with Mission Control and ended up, as far as we know, drifting in galactic space. 

Most markets outside China are more stable, so projections, even when they are wrong, don’t diverge quite so much from the actual situation.  Car sales are a great example. They are booming in China. Everyone I meet in Shenzhen, across all social classes, either has a car, is taking driving lessons or plans to begin soon. GM just announced its car sales in July in China rose 77% from a year earlier. 

For several months this year, China has been the world’s largest car market, outpacing the US. A quick web search turns up a supposedly highly credible forecast, from 2008, claiming that China is “on track to become the world’s largest car market by 2020, according to J.D. Power.” In other words, J.D. Power said it would take 12 years. It didn’t even take two. 

The recession in the US is a contributing factor, of course. But, the forecasts also, quite obviously, guessed very wrong about the growth rate of auto sales in China.   These wrong guesses have real-world consequences, because they can impact today’s decisions on investment and employment. In our case, by underestimating the growth rate of auto sales over the last two years, the projected revenues over 25 years from a $300mn toll expressway project in China also come in much lower. How much lower is anyone’s guess. Mine is that the revenue projections are off by at least 80% over the 25 years, and that this particular project will generate a profit of over $2 billion over that time, rather than the $1.2bn in the forecast built on the Hong Kong market researcher’s two year-old guesses. If so, the annual return on investment goes from the outstanding  to stratospheric. 

Here are my two projections: despite a record often unblemished by success, market forecasters in China will continue to ply their particular craft, collect their fees, sell their reports, and mainly miss the mark. Meanwhile, markets in China will continue to grow very fast, for a very long time.