Gambles Don't Belong in Investment

Gambling is very different to investing. Yet some people use questions about gambling to measure a person's financial risk tolerance and determine appropriate investments. Unfortunately, it's a flawed approach that is putting both clients and advisers at unnecessary risk.

'Gambles' is an unproven methodology for risk tolerance, with no scientific evidence to support it. It's unreliable, often returning different results when people retest. It's also a hard test for most people to do, containing complicated mathematics including probabilities.
 
But the problem with gambles runs deeper than the fact that they do a very poor job of measuring risk tolerance. At the same time as they are failing at their core task, gambles tests stand out as simply inappropriate in an investment advisory setting.
 
Professional money managers never have (and never will) promote themselves as gamblers. Their promise to investors is not that they are 'playing the odds' - it is that they are executing an investment strategy that is well conceptualized, properly researched and disciplined.
 
All of the mahogany on the walls and deep blues in the logos are about reassuring investors that they are professional, reliable, reasoned and dependable. All of those qualities are denigrated and diminished when complex investment decisions are seen to be made using questions about a coin-toss.
 
Meanwhile, people most likely think and respond differently to choices involving hundreds of thousands of dollars in retirement savings than they would for an idle $20 being put down on a roulette table.
 
Gambles tests were developed in university laboratories, using tiny amounts of money (often just a few dollars) where wins or losses really didn't matter. Most of the people taking the tests were students aged 18-21. It was not a representative population, or environment.
 
Critically, these questions were not being asked to help understand risk tolerance.
 
The people who came up with 'Prospect Theory' never had risk tolerance in mind - they were developing insights to help marketers overcome people's reluctance to make purchases because of the fear of the purchase not meeting expectations (a 'loss'). To this day, the creators of prospect theory have never supported claims that it means anything at all in an investment setting.
 
If the gambles methodology actually worked better than all other methods it might, perhaps, be worth 'working-around' the wrong messages gambles-tests send and their inappropriateness around investments. But there is no scientific data to say gambles work at all. All that pain is coming for no gain.
 
That's in stark contrast psychometric testing, which is the gold-standard for measuring financial risk tolerance. It has been rigorously tested in academic and real-life settings and proven to be accurate, reliable and dependable.
 
The science of psychometrics is well established and proven - psychometric testing works. It's been used to test financial risk tolerance for more than two decades, with millions of datasets collected.
 
The risk for investors who meet a gambles-based risk tolerance test is that it fails to accurately measure their risk tolerance, and they end up invested in inappropriate assets. If nothing goes wrong, they might 'get away with' this arrangement. But if things do go wrong, say a market correction, holding the wrong assets can spark a chain of financially devastating events.
 
The risk for advisers using gambles-based tests is that the finger of blame will be pointed at them by those financially-devastated investors. The regulatory, legal, financial and reputational risks of being found to have 'got it wrong' are very real - and very consequential. Today, even multi-billion dollar companies can be almost destroyed overnight by revelations that they failed to care for their retail investors.
  
For an industry based on understanding and managing risks, an analysis of the risks and rewards of using gambles-based risk-tolerance tests seems appropriate.
 
The risks are clear, and present - it is an unproven and unsupported methodology that produces unreliable results. Meanwhile, it's terribly difficult to see any reward for that risk - gambles is an inferior methodology compared to the alternative, which is psychometric testing.

Posted: 25/02/2019 10:17:02 AM by FinaMetrica Pty Limited