Preparing for the unique investment challenges of 2021

A strategic asset allocation (SAA) is an investment strategy whereby the insurer sets target allocations for various asset classes and rebalances the portfolio periodically. This is the single most important decision a captive’s board or investment committee makes on behalf of the investment programme.

The asset allocation decision, as opposed to the selection of an investment manager, accounts for more than 90 percent of the programme’s investment return of the strategic time horizon (Figure 1).

The allocation of mixed asset classes should be designed to produce optimal investment performance: put simply, to achieve the highest return for the lowest risk. The target allocations are based on factors such as the investor’s risk tolerance, investment objectives and time horizon. The “strategic” time horizon is generally accepted to be longer term (exceeding a single business, economic or interest rate cycle) and typically implies a 10-year forward-looking period. Of course, underwriting history, accounting, regulatory and other considerations should be factored into the analysis.

Generally, small to mid-sized insurance companies do not utilise the analytic systems which optimise asset allocations to produce the best risk-adjusted returns. Too many captives simply revert to an 80 percent investment-grade bond and 20 percent equity portfolio mix for their investment programmes. An optimised investment programme will typically contain between three and eight low or uncorrelated asset classes.

The percentage of the market value of the overall investment programme for each asset class should be optimised to produce the best risk-adjusted returns. That is, an investment programme is allocated in a quantitative manner to produce the highest return, per unit of risk, within the captive’s risk tolerance level, investment objectives, accounting and regulatory environment.

How to optimise your captive’s asset allocation

SAA analytic programmes of actuarial design—those that correlate both assets and liabilities using enterprise risk management principles—are clearly the best for insurers. All methodologies are heavily dependent upon an asset class’s expected return, volatility and correlation research data (assumptions), which may be obtained from leading long-term capital markets forecasters.

While some investment managers may provide internally derived assumptions, more unbiased and independent sources, such as those from large, industry-leading research firms like BNY/Mellon, Invesco, Blackrock and JP Morgan, are generally preferred.

Alternatively, captive insurer boards and/or investment committees, particularly those that have some degree of investment expertise, may be able to select low or uncorrelated asset classes and blend them in proportion to produce a return level and volatility (risk level) that they are comfortable with. Additional returns typically require taking extra risk unless significant diversification benefits are present. Fully understanding those risk/return tradeoffs is very important.

Modern portfolio theory (MPT) provides ratios and statistics enabling the comparison of return per unit of risk between asset classes as well as between investment managers. For a do-it-yourself approach to making decisions, captive boards should consider selecting asset classes with superior Sharpe and Sortino ratios, up/down capture ratios, and lower standard deviation of return, to name just a few of the available statistical measurements of manager skill or asset class efficacy.

Strategic asset allocation implementation challenges

Once an optimised portfolio of various asset classes is chosen, the next issue for captives involves implementation: how and when to fund each asset class within the optimised allocation.

While asset allocation is the most important determinant of long-term returns, manager selection can be critical over shorter time periods. In making manager selections, captives all too often default to simplicity: they will try to find a single investment manager that can manage all the asset classes selected. This clearly requires a large compromise, since single managers are rarely the best candidates across all asset categories.

We would also dissuade captives from the typical method employed to evaluate investment manager candidates. Most captives simply review the managers’ last five-year performance data vs the benchmark index. Unfortunately, managers can beat their benchmarks simply by taking more risk than the benchmark, which is not an attractive scenario for insurance companies. A review without consideration of the risks embedded in the managers’ portfolios would be insufficient and may lead to chasing past returns. In such cases a top performing manager may underperform when the market changes, as their successful bets or strategies of the past few years are cycled out of favour.

Ideally, captives should be basing their manager selection on risk-adjusted performance, which is the indicator of the manager’s skill level. The analysis should look at how much return was produced per unit of risk assumed. Unfortunately, captives typically do not have access to manager performance databases such as PSN Enterprise, eVestment or Investment Metrix, which allow for a comparison of manager skill within a large universe of managers (3,000+).

Absent a performance database, boards and investment committee members should minimally seek MPT statistics from several manager candidates for comparative purposes. Once a select group of finalists has been determined, based upon the quantitative MPT statistic comparison, the captive can be more assured that they are choosing from among the top skilled managers. That smaller group of managers should then be asked to conduct presentations, for proper due diligence, enabling the board to evaluate the finalist manager candidates on a qualitative basis.

Another critical asset allocation implementation decision involves the use of either passive or active management, or more likely both. Captives often struggle to find active managers that truly earn their fees consistently, by outperforming the benchmark. In some asset classes, this has proved to be quite difficult in recent years.

Oddly, in certain asset classes—bonds for example—about 75 percent of active managers beat their benchmarks. Therefore, and in those situations, a manager beating its benchmark is not a good indication that the captive has hired a highly skilled manager.

Allocations to some asset classes may also be too small to interest top active managers, whose minimum account sizes might exceed the captive’s threshold. Therefore, implementation of the optimised asset allocation may include passive exchange-traded funds (ETFs). In such asset classes, a passive allocation may be the most economic and feasible way to access the desired risk/reward characteristics of the asset class.

In other asset classes, however, such as investment grade bonds, passive ETF investments are typically third quartile performers, meaning 75 percent of active managers beat their benchmarks. Alternatively, and by definition, a passive ETF will always underperform, lagging the benchmark by its fee over time.

Captive investing: 2021 and beyond

With 2020 being a tumultuous year from market and economic perspectives, we wanted to share its implications for insurance investors. This past year we witnessed the impact of ultra-easy monetary policy, compressing bond yields and flattening the yield curve.

While the fall in interest rates created a not-to-be-repeated bond return windfall early in the year, the new near-zero interest rate environment, and building inflationary pressures, will make it difficult for insurers to achieve current income/investment cash flow goals going forward. With rates at zero and inflationary trends rising, it will be difficult for captives to produce real returns—meaning returns after inflation—that are positive. Under such circumstances, the most basic investment objective—that of preservation of principal—cannot be met.

To compound the issue, this zero-rate monetary policy was accompanied by an unprecedented fiscal policy which included massive business loans that were provided, and then forgiven, along with monetary hand-outs and unemployment insurance increases to soften the pandemic’s blow to the economy, businesses and citizens. Recent fiscal policy actions will also trend inflation higher over time.

Such massive manipulation of normal economic forces has caused equally large distortions in market and asset class behaviors. Figure 2 shows how the recent boosting of equity valuations due to fiscal and monetary policy has served to reduce future return expectations for investors.

As indicated above, curved efficient frontier lines demonstrate significantly lower expected returns for balanced portfolios, at any given risk tolerance level, than in the past. In effect, policy efforts to moderate the pandemic’s impact have simply “flattened the curve”. While the applied monetary and fiscal policies spared us from harsher consequences today, they simply blended the negative impact out over several years into the future. This is expressed in lower nominal and risk-adjusted return expectations for the same range of risk tolerance (volatility).

Table 1 clearly demonstrates the lower expected returns, per unit of risk, year over year.

In summary, the approaching investment environment makes it crucial for insurance companies to reexamine and reposition their investment programmes. What worked in the past simply will not perform in our new investment environment, in which the bulk of the captive investment programme—investment-grade bonds—produces a negative real return.

The captive board’s analysis should include a reevaluation of their investment objectives and risk tolerance as well as a determination of the most efficient path to investment success through asset allocation optimisation.

Carl Terzer is principal at CapVisor Associates. He can be contacted at:  carl.terzer@capvisorassociates.com