6 steps to more responsive asset allocation

By Mark Yamada | March 18, 2016 | Last updated on September 21, 2023
4 min read

This is the second in a two-part series that examines what the investment industry can learn from autonomous automobile technology. Read part 1

Self-driving cars must first sense, identify and avoid risk in their vicinity. Their second duty is to guide passengers to their destinations.

Last month, we compared two strategic portfolio approaches that address risk avoidance: constant risk (keeping a portfolio’s standard deviation constant) and fixed asset mix (rebalancing to 60% equities, 40% bonds).

From March 2002 to December 2015, constant risk demonstrated lower risk, lost less money during the worst 12 months in the period, and returned 11.8% annualized, versus 6.7% for a fixed mix. Conclusion: keeping risk constant and changing the asset allocation of a portfolio was more effective in protecting against changing market risk than continuously rebalancing to a fixed mix.

This month, we examine how to get investors to their destination.

Getting there

The GPS that guides us to our destination is familiar technology to most drivers. Newer systems route us around traffic congestion, warn us of problems along the way and allow us to make informed route choices.

A GPS receiver positions a vehicle using geometric trilateration from four or more of 24 operating global satellites. Further, GPS-enabled maps can monitor distance, time and speed travelled; estimate time to destination; and suggest route alternatives in poor conditions. Current portfolio navigation is embarrassingly naive by comparison.

Goals-based investing today

A traditional way to manage a goals-based portfolio is to have a separate portfolio for each goal (look at our case study in Advisor’s Edge, February). A modified method (see Figure 1) weighs the asset allocation of each goal by the investor’s priorities to derive a combined portfolio weight. The result is a single asset allocation that addresses all goals. The benefit is that investors see a connection between their objectives and the asset allocation. However, there are two problems.

  • A disconnect exists between the reality of future returns and the historical returns upon which models are dependent. History tells us little about future returns—monthly autocorrelation on the S&P 500 is only 0.03—and guarantees nothing (see “What is autocorrelation,” this page).
  • Rebalancing to a fixed asset allocation, while simple to understand and explain, is less effective than rebalancing to constant risk when it comes to managing risk.

Goals-based investing tomorrow

Taking a cue from GPS systems, we can apply trilateration to portfolios. Progress towards a client’s goal becomes the primary focus and the starting point is where we are today. For an investor looking to accumulate $2 million for retirement, for example, a capital accumulation path can be established (see Figure 2). We can manage the portfolio to the path. Actual experience will be above or below the path but, like a GPS receiver, the portfolio can be dynamically redirected to its path by increasing risk if the client falls behind (like A), and reducing risk if the client gets ahead (like B). The investor takes the returns the market gives her.

Better benchmarks

The capital accumulation path is a better guide to goal attainment than conventional index-based benchmarks, so measuring how far the portfolio is from this path is more useful. Conventional benchmarking uses asset weights expected to get investors to their destinations, informing investors and their advisors about comparative risk and performance along the way. But these benchmarks are based on observations gleaned through the rear-view mirror. Consistently beating the benchmark does little for the investor targeting $2 million if she ends up $500,000 short. How often have we heard that you can’t eat relative returns?

What is Autocorrelation?

Autocorrelation describes the degree to which the past influences the future.

Let’s say that, today, the market is up 6.6% over last month. An autocorrelation of 0.03 suggests that the market could be up or down any amount next month; past returns don’t tell anything about the future. Let’s say that VIX, however, was at 22.3. It has an autocorrelation of about 0.67, meaning that there is a reasonable chance that VIX will be in the vicinity of 22.3 in the next month (its daily autocorrelation is about 0.997, meaning there is a very good chance that VIX will be close to 22.3 tomorrow).

Defined benefit pension plans measure progress by tracking the difference between their assets and liabilities. Their funded ratio is similar to the distance from A and B to our capital accumulation path. This is a better way for investors to assess how they are doing. Portfolio risk can be redefined as the standard deviation around this path.


Using historical data, we tried to accumulate sufficient pension assets to replace 70% of income in retirement. In testing, the risk-on/risk-off approach achieved the goal 85% of the time, versus 37% for the 60/40 balanced fund. And, it did so with half the risk (standard deviation) in the critical five years before retirement (5.3% vs. 10.7%). Adding constant risk improved the likelihood to 97% (we published this study in the Rotman International Journal of Pension Management in 2011).

The autonomous automobile senses, identifies and avoids risks around the vehicle and automatically guides it to its destination. Portfolios can achieve similar results by following these six steps.

  • Establish a dollar goal and the path to get there (like a GPS).
  • Select a strategic asset mix that addresses the investor’s risk tolerance (a cruising speed), but recognize that it may limit their ability to get to their goal, and you’ll need to adjust it periodically.
  • Equate the strategic asset mix to a risk or value-at-risk number and rebalance to that, not to the asset mix (just like varying speed based on traffic and road conditions).
  • Use the capital accumulation path as the benchmark (again, like a GPS).
  • Add risk when the portfolio falls behind by an amount you pre-establish (like accelerating on clear roads).
  • Reduce risk when the portfolio is ahead (like slowing down in poor weather).
Mark Yamada headshot

Mark Yamada

Mark Yamada is president of PÜR Investing Inc., a software development firm specializing in risk management and defined contribution pension strategies.