Behavioral finance, the idea that psychological biases make investors less than the perfectly rational beings assumed by modern economic theories, continues to gain currency. Behavioral economists, in fact, have won two of the Nobel Memorial Prizes in Economic Sciences awarded over the past six years. PIMCO incorporates behavioral finance into its investment and risk management processes – particularly the finding that the psychic pain of losses can lead investors to make poor risk/return decisions. We believe insights such as these have grown more important amid high valuations for many assets and the maturing economic expansion.

For these reasons and building on our longstanding practice of cultivating diverse perspectives and challenging our own assumptions, PIMCO is partnering with the Center for Decision Research at the University of Chicago Booth School of Business to support the newly renamed PIMCO Decision Research Laboratories. Through this partnership, we hope to nurture insights into decision-making behavior that will ultimately help PIMCO make wiser decisions on behalf of our clients. The Center for Decision Research explores a wide range of topics in behavioral science, including behavioral finance and economics.

Now, let’s discuss loss aversion.

Quantifying loss-aversion bias

Daniel Kahneman, a 2002 Nobel laureate and early champion of behavioral economics, and colleague Amos Tversky were the first to document that most of us feel the pain of a loss more than the joy of an equivalent gain – a trait they dubbed “loss aversion.” They and others found that people need to “win” twice as much as they “lose” to be indifferent to taking risk.

An awareness of loss aversion requires us to rethink the familiar normal distribution, a simple image of symmetric curves showing equal probability of gain or loss. Understood in the context of loss aversion, the normal distribution is misleading because the response of risk-takers to losses is far more severe than the response to gains of similar size. Kahneman and Tversky’s loss-aversion ratio implies that in the mind’s eye, the mass of the distribution is approximately two-thirds to the left of the mean, and one-third to the right of the mean. The 2–1 relationship is a general rule that can vary by individual, the scale of the potential loss and whether or not participants are asked to “think like a trader” (in which case they become less loss-averse).

PIMCO’s approach

The concept of loss aversion is deeply embedded in PIMCO’s investment process, notably our stress testing and analysis of portfolio performance:

Ex ante loss-aversion management addresses potential losses in a trade or portfolio. The focus must be explicitly on the left side of the distribution. Quantifying tracking error, potential gain or expected return is insufficient. We must understand, “How much loss can be tolerated with this trade?”

Our stress testing draws attention to potential portfolio losses. We develop hypothetical forward-looking stress scenarios that represent adverse market outcomes for investment strategies. Risk budgets consider the size of possible losses in the event of an adverse scenario. This approach allows risk utilization to be scaled in the context of tolerable downside, with an emphasis on the sensitive left side of the distribution.

Ex post loss-aversion management entails two components of performance analysis: 1) measuring and 2) reminding.

Measuring compares how a drawdown differs from expectations and can be expressed in probabilistic terms. Consider a portfolio that has been running a tracking error of 250 basis points annually yet has fallen by 50 basis points over the past month. The decline is less than one standard deviation – a decline on that scale should occur more than 17% of the time. Quantifying an event in this manner enables an investor to determine whether it is sufficiently improbable to justify further attention.

Reminding can be more difficult. It entails looking at a portfolio and affirming that its risk profile was considered tolerable and appropriately compensated at the time of the initial investment. It may require an explicit act to overcome loss-aversion bias — bridging the gap between the forward-looking hypothetical willingness to accept a potential loss and living through a drawdown in real time.

Certainly, if a trade or portfolio suffers a loss greater than reasonably expected, deeper examination is warranted. In such cases, initial questions may include: “Was ex ante risk estimated correctly?” Or, “Is the market behaving unusually due to external events?” Answers then lead to consideration of actions, including reducing exposure due to uncertainty, increasing exposure due to more attractive value, or leaving exposure unchanged. Incorporating an awareness of loss aversion throughout the investment process allows us to maintain consistent framing on risk tolerance, leading to more objective, unbiased decision-making.

Loss-aversion bias may not be irrational

In practice (as opposed to theory), loss aversion may be not entirely irrational. Losses can create difficulties that are asymmetrical to the advantages that arise from gains. Investors, for instance, may feel the pain of loss more acutely because they recognize the difficulty of recovering from substantial drawdowns and the necessity to adjust exposure due to loss. In simplistic terms: The deeper you fall into a hole, the harder it is to climb out.

Numbers illustrate this point. Suppose a hypothetical portfolio of $100 declines 10% to $90. To return to parity an investor would need to earn 11%, not 10%. As further drawdowns are realized, a greater positive return must be earned to recover from the loss. A portfolio must generate a return of 100% to overcome a loss of 50%, for example. With a larger loss, recovery becomes far more unlikely.

There are other practical concerns associated with losses that investors may intuit. Negative returns in portfolios using derivatives can create margin cash demands, which, in turn, may require sales of holdings that the portfolio manager believes are attractive and would prefer to hold. Portfolio activity motivated by loss management is certainly less pleasant than trading to realize gains. In the extreme, highly leveraged portfolios may be at risk of such severe performance drawdown that the strategy must be entirely unwound. Mandate risk – the concern that underperformance might lead to an advisor being terminated as manager of a strategy – could contribute to a loss-averse risk profile in a portfolio. All of these circumstances represent potential outcomes indicating loss-aversion bias is not irrational.

An appreciation for loss aversion is arguably more relevant now as we experience bouts of market volatility amid trade tensions, moderately higher interest rates, and a global economy still growing, but slowing. Also, by incorporating these and yet-to-be-discovered insights from behavioral finance into our investment process, we aim to construct and manage portfolios ever more attuned and tailored to client needs.