the reason for the pop in life insurers from a rate rise is because their liabilities with be reduced due to the discounting and the increase in yields on assets supporting those liabilities…
Why would the rate change hit the liabilities and the assets differently? Aren’t they supposed to be more or less duration-matched? Are insurers loaded up on floating rate stuff?
(I’m not being sarcastic, I’m genuinely curious about this)
Well, he seems to be implying that the assets and liabilities are *not* duration matched.
That is such a general statement though, you can apply that to any market!
Yes, that is the implication. I just want to be sure that that’s the implication that was intended and not something else.
the problem is especially acute for life insurers which have durations that are “unknown” and normally go to 30 years…also considering the fact many insurers underpriced in the bull market as well…
so no, its not duration matched…keep in mind there are mortality assumptions built in which are surprising highly unreliable as well…
Theoretically, there must be a way to hedge these mortality assumptions… life insurance basically says “If I die early, pay me money”. There must be a way to make a contract that says “If I die early, I will pay you money”. If insurers do like 1,000,000 of these contracts, that should offset broad trends in life span at least.
Wow, check that out. Financial innovation right there!
Yes but are the uncertainties in mortality correlated to interest rate changes? Because that is what one would need to see for that to be relevant in an interest rate play.
So the argument presumably is that liability durations can go to 30 years, but there aren’t that many fixed income assets that can do that. There are some 30 y bonds now, stocks can have durations that high, and there are also a few perpetuities. However, assuming that these are hard to find, the assets are more likely to have durations of 8-10 years, while liabilities have closer to 30y.
in that case, I could see the rationale for a pop if interest rates rise. However, that would also imply that life insurance companies have been getting slaughtered throughout the 25y bond market bull run as interest rates go down. That doesn’t seem likely, although I don’t have the data in front of me.
Best to hedge life insurance with life annuities. In one, you pay more if people live longer; in the other, you pay more if people live shorter.
In general, life expectancy has been extending, so that tends to be good for life insurers and bad for annuity issuers. Of course, with chemicals and obesity and all that stuff, things can change.
Hmm. That’s a good point. I think one difference between annuities and life insurance, though, is age of customers. Annuities are mostly for old people. Life insurance customers are younger on average. So, annuity payoff will be more sensitive to general changes in life expectancy. I would expect that life insurance is more sensitive to accidents, or sudden manifestation of health problems.
Now, if they could get the old people to buy both annuities and life insurance, that would be pretty incredible…
its not really an argument…just read an annual report on the lifecos and you will see them discussing the interest rate assumptions and the impact it will have on their net income and balance sheet…i think its well known in the industry that higher interest rates will lift lifeco and p&c net income…and hence their stock
another factor hurting some lifecos are annualties tied to equity markets…
in terms of hedging mortality risk…who is going to take the risk ? they tried doing so but the market just isn’t big or developed enough…remember, insurers are there to take mortality risk to begin with similar to how banks take credit risk, you can offload it to somebody but eventually it ends up back on your balance sheet via securities or derivatives
For most life insurance companies, they can just handle the rise in interest rates by not duration matching…and avoid too much of a downside. BUT. Some of these insurance firms that actually have gifted investors, many of them have their assets in equities instead of short duration FI…they should come out best? People like Berkshire, Markel, WTM, Loews etc.
those guys you named are in P&C insurance which does not see as much impact from interest rates due to shorter duration of contracts and the ability to reprice…life insurance contracts are normally 20+ years in nature and does not reprice…duration matching is more of a concept for banks/p&c insurers…problem with insurance is that you don’t know when accidents/deaths will occur so your cash outflow is not necessarily predictable in practice…
if you wrote me a life policy, what would you invest in and how much would you charge? that is the question insurers ask themselves…
there are the law of large numbers but for some strange reasons lifecos get it wrong all the time…japanese lifecos went backrupt in the early 90s after japanese gov’t cut rates to zero…same phenomenon…certain industries benefit from higher interest rates (users generally include insurance, banks etc) and others get slaughtered by it (REITs)…
Are you going to coattail Berkowitz into AIG then? I looked at their balance sheet and gave up. Not man enuf for that.
This is a really good discussion. I think Frank and Palantir are really addding a lot here. Thanks.
Palantir brings up a good point that the duration risk is assymetrical, so they simply take some duration risk and figure it will ka-ching eventually, and for the meantime, companies will just use the premiums to stay cash flow positive.
Frank brings up the point that liabilities are hard to predict because mortalities can spike from time to time. That basically justifies a risk premium which is how insurance companies will set their premiums.
Rates getting cut to zero should hurt insurance companies if the logic you outlined is correct. But I’m not sure exactly how the Japanese experience shows that that lifecos are getting the law of large numbers wrong. It sounds like they expected mean reversion, took duration risk, and didn’t get it. That’s different from the law of large numbers not working out. The law of large numbers just means that mortality rates should converge to their age over time, and barring an epidemic or war or something, that is probably more reliable than interest rate mean reversion.
i stay away from lifecos but have bought p&c (way easier to analyze)…the accounting is basically a set of assumptions…in buying insurance cos in general you really have to have a sense of the risk mangers appetite…just cause something is priced under book isn’t saying a lot when certain liabilities are kept off due to aggressive assumptions…this is why i like BRK, Lowes etc cause i understanding the thought process of value guys (minimize risk, ability to say no) etc…