Is it true that “the easy money has been made?”
This is something that it has been common to hear over the last couple of years and, it seems, is even more common today.
But, as Morgan Housel pointed out in 2015:
A similar argument was in a piece from last year (which also provides a nice summary of some of the major market episodes of the last decade), so I’m not going to repeat the points made there. But I believe there are some observations worth making today.
What was ‘easy’ about it?
“The easy money has been made” is one of those clichés which gets bandied around but doesn’t tend to merit closer inspection. We think we know what it means, but the phrase can convey a number of different arguments, and often a mixture of them all.
Do we mean that the returns generated were ‘easy’ in the sense that they were ‘foreseeable?’ Or were they ‘easy’ as in ‘not emotionally challenging?’ Or do we mean that it was easy because a wide range of assets have done well, so you haven’t had to work too hard on selection? Or, more commonly these days, do we simply associate ‘easy money’ with ‘easy policy?’
Thinking about what we mean by ‘easy money’ can reveal something about how our own biases work, the sentiment of markets today, and how we should consider valuation assessments.
Reinventing history: hindsight bias
Changing investor perceptions about risk may not just be about greater comfort about the future, but can also involve a reappraisal of the past.
As human beings we have a tendency to forget how we used to feel. One of the most commonly known manifestations of this is ‘hindsight bias’: we tend to interpret the progress of past events as more foreseeable than they actually were (‘I had a feeling that was going to happen’). It is an ego defence mechanism: the brain doesn’t want to confront us with all the times we were wrong so it shields us from our errors wherever possible.
Because we have seen strong investment returns, we look back for signs along the way to explain why it happened. By focusing on these at the expense of other signals, we think that the rally was more obvious than it actually seemed (as is often the case, the phrases we use in our day-to-day lives often reveal our biases, consider the sentiment: ‘all the signs were there’).
Path dependency and the ‘peak-end’ rule
A less well-known behavioural trait is the ‘peak-end rule.’ Psychologists suggest that the way we evaluate the past (for example as a ‘good’ or ‘bad’ experience) is based not on how we felt over the whole journey, but by considering single points of time along the way.
More specifically there is evidence that we overweight our more recent experience (the ‘end’) and the most extreme moments (the ‘peak’) and tend to neglect the how long we spent in a specific state. In trials for example, participants were subjected to two conditions:
- spending a minute with their hand in uncomfortably cold water followed by an extra 30 seconds in slightly warmer (but still cold) water,
- or just the minute in the cold water
When asked which experience they’d prefer to repeat, most chose the first scenario even though you’d think that was a worse experience overall. Because they remembered the (slightly) more pleasant ‘end’ of the first scenario they evaluated it more favourably than the constant pain of the second. They paid less consideration to the length of time spent in each state.
(This may seem less in tune with our own experience than hindsight bias, but consider the occasions when we use phrases like ‘that wasn’t so bad,’ or our decisions to repeat a challenging experience even after promising ourselves ‘never again’ while we were doing it first time around).
Path dependency therefore matters for how we think about the past. Consider stock markets over the last five years:
Our recent experience has been very good. It makes us feel better about the peak pain at the start of 2016, and certainly causes us to neglect the period of close to two years where we had made no gains in price terms. We feel different about where we are today because of the route we took.
Another example of the ‘easy money’ argument may be that: “it wasn’t predictable that we would get the returns we did at the time but if you had told us ahead of time what policy rates would do and how much QE there would be, then the subsequent returns were obvious.”
There may be something in this, but it isn’t a useful observation for investors. Policy rates play a critical role in anchoring the valuations of all assets but not in a linear way. For example much of the last decade has not been associated with equity markets re-rating in line with falling interest rates – quite the opposite in fact.
Risk premia were also moving because the very environment which engendered supportive policy was one in which we spent most of the time in fear that systemic collapse or recession were just around the corner.
Nor can we say that a tightening of policy will mean a symmetrical reversal of asset price behaviour while policy was easing. As has been the case over the last year, rising policy rates can be associated with strong growth and profits outcomes. It may also be that interest rates do not rise. This would likely mean that expected returns over the long term are lower than we have been used to, as Anti Ilmanen recently outlined.
The ‘era of easy money’ was actually one of deep fear and discomfort. This is exactly why assets were offering the compensation for risk that allowed strong returns to be generated.
It is true the declining real interest rates around the world played an important role in this (though it is perhaps odd to consider it an era of ‘easy money’ when cash rates are negative after inflation) and the path of rates from here is crucial. If rates do not move, and other assets re-rate further, it may also mean that we reach a point where the level of returns generated may not be repeatable.
However, if the end of ‘easy money’ means an end to the fear of the last ten years and higher returns on cash and other less volatile assets, then maybe it should be welcomed.