With billions of dollars worth of variable annuity contracts with in-the-money living benefit riders leaving their surrender periods over the next 18 months, the issuers of those contracts will soon learn how accurate their lapse rate assumptions have been.
For those who guessed right, no great shocks are likely to occur. But if issuers erred on the high side, say analysts at Oliver Wyman Group, lower-than-expected lapse rates could translate into higher-than-expected reserve requirements and a potential dent in profitability.
Generally, the presence of income guarantees is expected to result in stickier policies. “We should see lower lapses than in previous nonguaranteed blocks,” said Todd Solash, partner, and senior manager Aaron Sarfatti at the global consulting firm, which advises insurance companies on risk management and overall strategy.
As a result, aggregate reserve requirements could rise by $1 billion to $5 billion over the next 18 months for the $400 billion in outstanding VAs with living benefits, they said.
“This is approximately 10% of total industry general account reserves for living benefit guarantees,” says a recent Oliver Wyman report, A question of legacy: Measuring and managing behavioral risk in variable annuities. “Should this adverse scenario unfold, it will slowly, but surely, erode the capital position of many leading variable annuity writers who would have to post reserves and capital for the policies they had expected to lapse.”
Most issuers did a good job of hedging the market risk associated with their VA guarantees, but they didn’t necessarily anticipate policyholder behavior risk in their models, according to the report. “The magnitude of the behavioral risk wasn’t as well studied as market risk,” Solash said.
The potential magnitude of lapse rate risk is still unknown. According to the report, if the S&P 500, now at about 1,050, rises to 1,250, the reserve requirements will be much lower than if the index drops to 800. The impact will also vary by carrier, depending on the size of their VA/GLWB book of business, their lapse assumptions, the degree of “in-the-moneyness” of their riders, and the channel in which they distributed the contracts.
About 14% of issuers believe that lapse rates vary by channel, according to a 2009 Society of Actuaries survey of 29 issuers. “All of those indicating a difference distinguished between internal distribution (captive agency) and external distribution (brokers, banks, independent agents), with the latter having higher lapses,” an SOA report noted.
Policyholder persistency would be good if most variable annuity guarantees were out-of-the-money or at-the-money. Many contracts are in the money, however, despite the partial recovery of equity markets from the 2008-2009 market crash.
Issuers would be happy to see those contracts lapse. But contract owners are less likely to want to give up the now-valuable guarantees. Their advisers, mindful of suitability standards, are also less likely to recommend 1035 exchanges that would nullify the guarantees.
Therefore lapse rates in coming months might be lower than the actuarial assumptions. “[Many carriers] used historical 20% lapse rates, and the guarantee will reduce the lapse rate. It took the crisis to test the assumptions,” said Sarfatti.
The deferral bonuses or “roll-ups” that carriers offered to discourage early withdrawals from VAs with income benefits are also likely to make the contracts stickier. “I wouldn’t call [the rollup] a helpful feature” in this context, Solash told RIJ.
“The rollups can be risk mitigating in product design, but with falling interest rates and equity prices, roll-ups greater than the interest rates can be a very expensive feature.” He doesn’t believe that roll-ups will be a primary cause of lapse rate surprises, however.
The “rationality” of policyholders is one of the wild cards in forecasting lapse rates. A perfectly rational policyholder (such as a professional investor or hedge fund manager) might be counted on to surrender an out-of-the-money contract and keep an in-the-money contract. But policyholders aren’t always rational, and don’t necessarily take maximum advantage of their guarantee. Some are forced by economic hardship to cash out their contracts at the end of the surrender period.
People have also presumably purchased VA/GLWB contracts for a variety of reasons. For some, the income feature may have been the primary reason. For others, it may have been a secondary consideration. For those clients, lapse rates might be higher.
Along with the Society of Actuaries, several consulting groups have studied VA/GLWB lapse rates in recent years. Towers Perrin, Deloitte LLP, and Germany’s Risklab are among them.
They tend to agree that there’s a shortage of hard data on behavioral risk. “Historic experience is sparse for variable annuities, and companies usually have a very high level estimate regarding the level of rationality of their policyholders,” said Deloitte actuaries in a slide presentation, “Residual and Ancillary Risks in VA Guarantees.”
In its research paper, Oliver Wyman suggests that in the future VA issuers take the following steps to protect themselves against behavioral risk:
- Improve behavior modeling and forecasting capabilities
- Integrate behavioral risks more fully into risk management protocols
- Revive the variable annuity reinsurance market
The paper also suggested that VA issuers answer these questions:
- Which of my in-force blocks are most exposed to behavioral risks?
- How conservative are my lapse and utilization assumptions relative to peers? Relative to experience?
- What would be the impact on earnings and capital through 2012 if lapse rates for guaranteed businesses fall 25% and 50% below expectation?
- How does market performance affect this impactWhat is the transaction cost incurred via the dynamic hedge program? By how much would this expenditure decrease if market exposure tolerances were doubled?
- More broadly, how should I incorporate the uncertainty around behavior in the design and execution of my hedging strategies?
- Do I have a business plan if behavior is worse than expected over the coming years?
