Probability is power but numbers alone can’t change behaviour


Craig McCulloch

Uncertainty is a fundamental part of investing. The outcome of fluctuations in economies, markets, asset classes, regulations and investor behaviour is never more than a probability.

Monte Carlo simulation can empower investors to understand such fluctuations like never before.  However probabilistic information can also make many prospective retirees uncomfortable.

However high a model might suggest, a probability is no guarantee and retirement is a one-stop destination. Make the wrong decision and there may be no way to make up for a plunge in value of a lifetime of savings.

For example, a retiree may have a goal of generating annual income of $50,000 over the next decade and have a 95% chance of achieving it by following a particular financial plan. That is a high level of certainty, but cold comfort if the outcome ends in failure as was predicted to happen 5% of the time.

Such low probability events do occur but those unfortunate retirees aren’t aware of, and care little for, the hypothetical scenarios that did predict they would achieve that goal. In fact, they’re likely to label the initial probability forecast and financial plan “wrong”.

Simple probabilities, complex discussions

Financial planners need to be finely tuned to these issues. They need to communicate future uncertainty in a nuanced and careful manner, teasing out the underlying assumptions that investors make.

The probability of achieving an initial goal is a binary outcome: succeed or fail. It does nothing to quantify the magnitude of “failure”.

In the previous example, an investor arguably needs to know more about how far short they might be in the event they don’t meet their $50,000 annual income target, and plan accordingly. Is the miss likely to be by one year or by several years? Are they likely to miss their target by $1000 or by $20,000 a year? The comfort levels of an investor will naturally differ greatly between these two scenarios.

However, this is just the beginning of the discussion. Investors are not rational beings who weight their decisions according to the probabilities of outcomes. They are prone to numerous behavioural biases, and advisers need to be aware of this and have a strategy to help guide wise choices.

Loss aversion is a powerful driver, particularly for retirees, who tend to be naturally more risk averse. It is a key reason why people overweight decisions when presented with improbable outcomes (the possibility effect) and why they underweight near certain outcomes (the certainty effect).

In practice, this means an investor with a 95% chance of achieving their target will be overly risk averse because they fear disappointment. A study by Daniel Kahneman and Amos Tversky (the authors of Prospect theory) found that people applied a decision weight of just 79.3% when given a probability of 95%. The corollary is that investors don’t fully appreciate the difference between, say, an 80% probability of an event and 95%.

Framing can also play a role. Investors are more likely to place emphasis on scenarios that are told vividly and emotionally, and low probability events are more heavily weighted when presented as a relative frequency (e.g. “five out of 100 retirees ran out of money during retirement” compared to “95% of retirees met their income goals”).

Better information powers great advice

The power of sophisticated analysis, combined with more effective communication from advisers taking into account behavioural biases, can help investors achieve their goals like never before.

However, an adviser has to ultimately make a judgement call on how much to educate and shift the natural position of clients and how much to push them. This comes down to the personal goals of clients.

There’s a natural tension between risk tolerance (what investors can withstand emotionally) and risk capacity (a measure of how much risk they should take in order to achieve their goals).

Monte Carlo analysis can help make the difference between the two clear. With this information, discussions between advisers and their clients can become significantly richer. For many investors, this might require taking on greater risk but balancing this with stronger risk management techniques.

Good advice is an ongoing service even in retirement. The uncertain nature of the future implies that ongoing adjustments are necessary, just as the goals of retirees also naturally change over time.

Probability doesn’t mean retirement has to be uncertain. It means that retirees have more than one shot at getting it right.

By Craig McCulloch, senior consultant 


You must be logged in to post or view comments.