If You’re So Smart, Why Aren’t You Rich?

booksThe title of this entry is a cliché, one either used or attributed to the wealthy uneducated at the expense of their college-trained but not-so-wealthy brethren. It is akin to asking, “How is all the book learning working out for you?”

For the academic economist, the jibe has a special point. For economists are thought to be experts precisely in the operation of markets: the one subject you might want to be an expert in if you were trying a priori to figure out a good way to get rich.

A Consoling Answer

The smart economists of recent decades, especially those who have turned their attention in a serious way to the quest for alpha, have often drawn the conclusion that alpha is an illusion. Alpha nets out to zero (by definition), and the effort to get a positive alpha at the expense of some else’s negative is doomed by the efficient nature of the markets on which it is to be attempted. This gives a plausible answer to the question above. “I’m too smart to engage in a fruitless pursuit of wealth – after all, I don’t buy lottery tickets either, though I concede some people do win.”

It is a fair answer. But it doesn’t stop the questions, and there are some market/real-world winners whose winning seems to have something more to it than that of the lucky lottery contestant. It would be fun to be able to say, “Some of us economists do get rich from the use of our erudite insights. Here are some names.”

And this brings us to a new paper by Thomas Raffinot, a senior market economist at BNP Paribas Cardiff. The title is straightforward, “Can Macroeconomists Get Rich Nowcasting Economic Turning Points?”

Windows underlines the word “nowcasting” with a red squiggle, indicating that it is still something of a neologism. Yet it is an obvious enough portmanteau of “now” and “forecasting,” and I presume MS will catch up with it in time.

Let the Robot do the Heavy Lifting

Raffinot celebrates the invention of Learning Vector Quantization (LVQ). This bit of taunted-academic’s revenge is a machine-learning algorithm. [“What good is book-larning, Uncle Jeb? Well heck, it’s brought us to where we can del’gate the task to machines!”). LVQ can find a natural grouping in a set of data.

Economists have recently assigned their LVQ robots the task of picking turning points in the growth cycle. Not the business cycle, I hasten to say, the growth cycle, which is the deviation between the GDP and the long-term trend, so that turning points separate periods of acceleration on the one hand from slowdowns on the other. This has straightforward investment/trading consequences.

One great thing about the growth cycle is that it is possible to follow it, effectively, in real time – hence the “now” in “nowcasting.” Business surveys “provide economists … timely and reliable pieces of information on business activity” that are published “before the end of the month they relate to or just a few days after.” In the U.S., examples include the surveys published by the Institute for Supply Management, the Conference Board, and the National Association of Home Builders.”

LVQ can also receive as input the weekly initial claims for unemployment insurance and high-yield corporate spreads.

Raffinot runs some tests, and concludes that LVQ “turns out to be very effective … and that some economic indicators can be exploited to quickly identify turning points in real time in the United States and in the euro area. It leads to useful implications for investors practicing active portfolio and risk management and for policy makers as tools to get early warning signals.”

So … an economic education may help you get rich after all.

Until such time, anyway, as the robots who run the LVQ algo figure out that they don’t need you and open up their own bank accounts.

 

 

 

 

 

Be Sociable, Share!

Leave A Reply

← The Quant Conundrum: Perfectly Repeatable or Repeatedly Imperfect? Let’s Hope Nothing Comes of Sprecher’s Grand Bargain →