Investing amid rising market uncertainty
In Germany, the largest economy in Europe, gross domestic product (GDP) growth fell into negative territory in Q2 2019, led by a slowdown in manufacturing, and the US Institute of Supply Management Manufacturing Purchasing Managers’ Index fell below 50 in both August and September, denoting contraction in manufacturing.
The global manufacturing slowdown has been exacerbated by continued trade policy uncertainties and dented business confidence resulting from geopolitical risks, including the UK’s departure from the EU and the imposition of trade tariffs between the US and China, and potentially Europe.
While manufacturing comprises only 11 percent of US GDP, there are risks that a manufacturing slowdown and continued uncertainties will further curtail business spending, eventually causing job losses. Rising unemployment, from a 50-year low, could dent US consumer confidence leading to falling consumption, ultimately tipping the US into recession (consumption represents 70 percent of US GDP).
Central banks including the Federal Reserve in the US, concerned about low inflation and rising global risks, have started to cut interest rates. With policy rates negative, in for example Europe and Japan, investors are questioning whether US rates could also become negative during the next recession.
Since January 2019, falling interest rates have increased values of captive insurance fixed income assets. Over the first three quarters of 2019, for example, the ICE Bank of America Merrill Lynch 1-5 year AAA-A US Corporate & Government Index returned an annualised 5.6 percent (compared to 1.5 percent in FY 2018), and US 10-year treasuries returned an annualised 14.8 percent (compared to 0 percent in FY 2018).
While fixed income asset values increased during this period, so did the present value of projected liabilities, which are typically discounted at risk-free rates. Thus, net of liabilities, it is possible that captive insurers would have been no better off.
In a low interest rate environment, captive insurers may be tempted to extend portfolio duration and/or invest in lower quality fixed income assets to achieve higher returns. However, this can be fraught with risk.
If a captive insurer’s fixed income assets are longer in duration than its liabilities, overall investment performance, net of liabilities, can be negative if interest rates rise—unless other investment income and/or assets, such as equities, can overcome the gap. Similarly, net investment performance can also be negative if interest rates fall and a captive insurer’s fixed income asset duration is short in comparison to its liabilities.
An asset liability management (ALM) framework, which compares interest rate sensitivities of assets vs the liabilities captives intend to match, is essential to managing interest rate risk—especially for insurers with longer-tailed and more interest rate-sensitive risks, for example medical malpractice and workers’ compensation-related liabilities.
Amid low yields, some insurers are purchasing high-yield fixed income assets, including private loans, in order to achieve higher returns. These assets, however, come with increased credit and liquidity risks. If these assets mature in full and on time to pay off expected liabilities, credit and liquidity risks may not be a major concern.
However, if projected liabilities are different in practice—for example, a new captive with limited claims experience may have trouble projecting liabilities—captive insurers could crystallise losses should they liquidate large portions of their portfolios prematurely.
Reviewing the numbers
Figures 1 to 4 provide an illustration. Figure 1 shows a hypothetical portfolio, consisting of 70 percent government bonds and 30 percent investment grade (IG) bonds, matching projected insurance liabilities of 100 in years one to five; all bonds successfully mature at par.
Figure 2 is similar to Figure 1, except that the portfolio holds 20 percent government bonds and 80 percent IG bonds―this portfolio reflects the captive insurer’s desire to generate higher investment income by investing in higher yielding assets.
In Figure 3, the captive insurer decides to buy higher yielding short-dated and long-dated assets and decides against matching projected liabilities. It is important to note that all three portfolios have the same duration of three years.
Figure 4 shows hypothetical government bond yields and IG credit spreads during normal and stressed environments. During a period of market stress, government bond yields drop approximately 25 basis points (bps) while IG spreads over US treasuries double causing five-year IG yields to surpass 3 percent (in reality, these levels could be more extreme).
In Figures 1 and 2, if liabilities occur as expected, then bond maturities will finance expected payouts perfectly. This is not the case in Figure 3, where there is reinvestment risk after the one-year bonds mature (ie, the surplus cash from the maturing bonds may not be invested at rates comparable to where liabilities are priced for years two, three and four)
During market stress, the portfolios in Figures 1 to 3 lose 0.65 percent, 1.74 percent, and 1.96 percent, respectively vs the value of projected liabilities (ie, after applying the scenarios in Figure 4). Should the captive insurer need to liquidate assets to finance unexpected liabilities, a portion of these losses could be crystallised, with the portfolio in Figure 3 suffering the greatest loss.
When compared to post-dividend combined ratios of 96 percent (leaving behind 4 percent for the captive insurer) in 2018 for AM Best-rated captive insurers, a 1.96 percent loss is sizeable.
To improve economic viability, captive insurers have explored ways to reduce claims and expenses related to liabilities while increasing the value of their assets. To reduce liabilities, insurers have explored flooring liability discount rates (eg, at 0 percent). However, when cash earns a negative rate, it is reasonable for the present value of a future liability to be relatively greater since, inflation aside, cash deposited today can lose value over time.
Thus flooring discount rates is not ideal. Instead, captives could explore other solutions such as reducing claims (eg, through promoting preventive medical wellness programmes) and reducing administrative costs.
With respect to assets, in addition to increasing corporate bond exposure, insurers can explore other options including: i) purchasing currency-hedged foreign government bonds, which can earn spreads over US treasuries; ii) buying inflation-linked bonds, which are currently inexpensive by historical standards; and/or iii) allowing for greater investment policy flexibility accommodating broader asset class inclusion.
When IG spreads over treasuries are low, other assets may provide better risk-adjusted returns. Regulators, too, can help by creating more accommodative frameworks that enable captives to take broader investment risks within appropriate risk management frameworks.
For captive insurers with limited claims experience and/or those facing policy uncertainty (eg, healthcare captives), it may be prudent to match portfolios closely against projected liabilities. As added protection, investment policy statements can mandate holding sufficient high quality liquid assets to help avoid deeply discounted asset sales during market stress.
With more claims experience, maturing captives may want to consider broadening their investment mandates to incorporate additional flexibility and asset classes.
With central banks and investment portfolios entering uncharted territory, it will be vital for captive insurers to be adequately prepared. l
The views and opinions expressed herein are those of the author and do not necessarily reflect those of Butterfield Bank (Cayman) Limited or The Bank of N.T. Butterfield & Son Limited.
Zafrin Nurmohamed is senior portfolio manager, asset management at Butterfield Bank (Cayman). He can be contacted at: email@example.com