Merrill released an updated research piece on Apollo which was insurance focused, and rightly so, given the >60% of earnings that are derived from annuities going forward.
Apollo has argued that its insurance earnings, or “spread earnings”, should trade closer to what more traditional asset management earnings streams trade at:
Respectfully, I don’t think that should be the case. What is missing on the left hand side of the slide are all of the complexities that are bundled with sourcing your capital from an annuity versus a traditional source (drawdown vehicle, REIT, BDC, etc.). It is, in my opinion, precisely why Blackstone won’t touch it with a ten foot pole.
Those complexities boil down to the simple fact that it’s not a cakewalk to “duration match” the assets with the annuity liabilities, and therefore one always has risk on the funding source side in addition to the investment side. The duration of the liability is a moving target, governed by surrenders, crediting rates, minimums, etc. Insurers in the 90s/00s got stuck in the mud within the seemingly “simple spread based insurance” business as rates went to zero and their modeling of lapse rates went out of whack (more people hung around than modeled, extending duration vs projection). Accordingly, they were left holding the bag reinvesting into a low yield world when their liabilities were higher yield.
Complexity increases as the attempted duration match involves “net duration matching,” meaning each liability will have a pile of assets that on a net basis, roughly equal the projected duration of that liability. Apollo’s business has liabilities with an average duration of just under 9 years. An insurance industry insider I chatted with posited that they think Apollo along with the rest of the Alts participating in insurance from an underwriting angle are not duration matching their liabilities when you peel back the onion. It’s not easy to originate assets with duration of 9 years or north of that. 5-7 year duration assets have a much larger pool to fish from. As such it’s possible that these insurers’ “net duration matching” is a bit less simple and secure that they indicate, perhaps including a mix of longer duration and shorter duration assets that don’t effectively always move in lockstep with the duration of the liability.
It further increases when one has to consider that a fixed indexed annuity is also marked up in line with a portion of the S&P500 return (the benchmark may vary), forcing the insurer to own a pile of derivatives to hedge the equity upside risk. While the hedging isn’t rocket science, it does have its own set of variables that change with implied equity volatility. And it is additive to the overall complexity of the insurance earnings stream.
Last, the cash flow of insurance streams is based on statutory releases of capital, which vary based on all the changing variables (and more) that are discussed above. They aren’t as simple as the management fee paid quarterly like clockwork. Real cash flows of insurance businesses are notoriously mismatched with perceived earnings, and for the most part, pushed into the future.
All is to say, to conclude, upon digging it should be clear that the complexity of insurance liabilities as one’s capital source is much, much higher than traditional capital pools (e.g., an endowment writing a check into a limited life vehicle with no ability to early redeem, ever). Asking the market to value this earnings stream closer to traditional asset management not entirely unreasonable, but a stretch. And when a company is going all in on the insurance side of their business versus the traditional side of the house, it may be appropriate to value the overall business at a lower multiple than its peers of the past.
Disclosure: We own shares of Apollo Global. This is not investment advice. Do your own work.