One major trend that has existed in the insurance industry over the past several decades is increased specialization. This is true globally, but is most true in the US. In general, being good at one area of the insurance business does not confer significant advantages in other areas. This is true in underwriting, which is a specialized talent in each area of insurance. It is also true in market. Cross-selling opportunities across lines of business are not great enough to justify the effort. There are small consolidation advantages in shared corporate staff, and it may be cheaper to finance a large organization than two smaller ones, but those are slim advantages to conglomerate over. In general, we do not expect an increase in the number of major conglomerates; rather, the trend is toward increasing specialization, with increasing scale within a company’s area of specialization.
The most successful conglomerates have tended to be holding companies that allow individual operating companies to act with little coordination. They require results from the subsidiaries, and intervene when the holding company does not get results. The holding company directs the allocation of capital to the businesses that offer the best prospective returns. In one sense, some insurance conglomerates invert the insurance model; they are essentially investment companies, with insurance liabilities issued to provide funds to invest in favored projects. With ordinary insurers liability issuance leads the process.
The excellent performance of the conglomerates group is unique to two companies that are large and do it particularly well: AIG and Berkshire Hathaway (BRK.A). As these companies get larger, it will be increasingly difficult for them to achieve above average performance. It will also be more difficult to do as well without their unique current leaders.
Bringing It to the Present
Well, when I wrote about this in 2004, I had not yet focused on the growth of debt inside AIG. The firm I was with owned AIG, and sold into the flood of buying that occured when AIG was added to the Dow.
But I can adjust my views. Before the time Greenberg was being shown the door, I was bearish on AIG because I thought that leverage was too high, far greater than the rating agencies should allow, which explained why AIG credit traded more like a single-A credit. I also reflected on my prior days at AIG with more insight than when I was younger, and wondered if it might not be possible that the accounting at AIG was very liberal. While working there, I was involved with uncovering five reserving errors, each of which should have led to a hiccup in earnings, and in one case, a severe loss. But the earnings kept going up.
When I have written things like this, I have never gotten flak from people inside AIG; rather, I get e-mails from people inside AIG that have had similar experiences. Part of the problem was the culture of fear. When telling the truth is initially derided with force, even if it is eventually accepted if you are strong enough, you will get few people taking the risk to tell the truth.
But without Greenberg, AIG was doomed in the short run. No one else could manage it, even if it was crooked in some ways. The truth withheld re-emerged, and those in charge got whupped for the sins of Greenberg.
But what of insurance conglomerates generally? I don’t recommend them, because synergies are few, and focus is necessary for most effective insurance companies.
Banks and the Insurance Business
My bias in understanding banks in the insurance business is that banking and insurance are fundamentally different businesses, but there are areas of overlap where the participation of banks sense. In Europe, indiscriminate mingling of the two businesses has usually led to losses. Why?
Though banking and insurance are both described to be financial services, they are different in the terms of financing done, arid service provided. Here are some of the key differences:
- Product complexity: Insurance liabilities are typically more complex than bank liabilities; there are more factors that can affect the overall cost of the promises that an insurer makes to a policyholder, than a bank makes to a depositor.
- As a result, the liabilities underwritten by an insurance company are usually riskier than those underwritten by a bank.
- Because of the relative riskiness of the asset and liability structures, including the greater length of guarantees made, insurance companies generally run at a higher ratio of book equity to assets.
- With the longer liability structures, and a highly competitive environment, the investment policy of most insurance companies is more aggressive than that of most banks. Interest rate risk is not generally a problem; most companies attempt to squeeze out interest rate risk by approximately matching assets and liabilities. Most of the risk comes from investing in equities, lower grade corporate debt, and equity risk from the writing annuities. (As the market rises and falls, so do fees received.)
- Liabilities are more expensive to originate and service at insurance companies.
- There is a high amount of idiosyncratic expense associated with running an insurance company. When a bank buys an insurance company, there are usually few expense savings.
- Though there are diversification advantages from a holding company owning both banks and insurers, this advantage often outweighed by the different skills needed to manage the different entities well.
The European experience with banks and insurance companies under the same roof has been mixed. One success has been banks coming to dominate distribution of life insurance products. Banks distribute more than 50% of all life insurance policies in most countries in developed continental Europe. The tendency for banks to sell insurance is strongest in countries where banks are dominant financial institutions aside from insurance.
But there have been failures as well. Most of the failures have been due to a lack of understanding of how different banking and insurance really are. Others have been due to taking too much risk, particularly in unfamiliar countries. Here are some examples:
- CSFB buying Winterthur did not grasp how sensitive the performance of Winterthur was to the performance of the equity markets. When the equity markets fell, CSFB had to pump in $2.4 billion of capital.
- Allianz (AZSEY.PK) did not grasp the poor asset quality of Dresdner, particularly in the midst of a bad market for investment banking
- Zurich Financial Services (ZFSVY.PK) was overly aggressive in the expansion plans in the US, leading them to overpay for marginal asset management companies like Kemper and Scudder.
- Aegon (AEG), ING (ING), and Prudential plc (PUK) all suffered by building up leverage through 2000, particularly in their US life insurance subsidiaries, and then got whacked by the combination of the bear markets in equity and credit.
To summarizing the European experience positively, if a bank has strong customer relationships, it can earn additional margins through distribution of insurance products. Negatively, conservatism pays in entering new lines of business and new countries.
The US experience with banking and insurance together has been more limited, due to laws such as McCarran-Ferguson and The Bank Holding Company Act. McCarran-Ferguson, passed in 1945, entrenched the exclusive authority of the states to regulate insurance. The Bank Holding Company Act of 1956, amended in 1970, restricted the insurance activities of bank holding companies.
Until HR 10 was passed in 1999, the Federal Reserve gradually relaxed regulations on bank involvement in insurance companies so long as earnings from insurance activities remained below a threshold. In April of 1998; the merger announced between Citicorp (C) and Travelers (TRV) forced the need for structural legal change, leading to the passage of HR 10, otherwise known Gramm-Leach-Bliley. HR 10 allowed for the formation of financial holding companies that could engage in banking, investment banking, and insurance, with regulation of mixed entities to be done functionally down at the operating companies, in much the same way it would be done for standalone entities.
When HR 10 was passed, there was a lot of expectation in the insurance industry that the new law would have no large effect. Some observers suggested that life and personal property/casualty insurers might be bought by banks because of investment and product marketing synergies. But most thought that banks would not buy insurance companies, and insurance companies would not buy banks. This expectation has largely been met. Aside from Citicorp, only Bank One has acquired an insurance underwriter of significant size.
Even with Citigroup (neé Citicorp) the acquisition of Travelers was re-thought. In 2002, Citigroup spun the property/casualty operations off as Travelers Property Casualty, which had a short-lived existence as a standalone company before merging with The St. Paul. Citigroup kept the Travelers life and investment operations (and the logo).
Bank One acquired the US life insurance operations of Zurich Financial Services. This allows Bank One, soon to be a part of JP Morgan Chase (JPM), to underwrite life insurance. They presently use it to sell term insurance and annuities.
So, why didn’t banks attempt to enter the life and personal property casualty lines, in general? The quick answer was that they didn’t need to; many already had the benefits that come from distribution of insurance products, without the additional risk of underwriting, the additional hassle of state regulation, and the complexities of managing two disparate businesses. Additionally, the sale of insurance products has tended to be a high ROE business, whereas underwriting, given the stiff capital and reserving requirements, tends to be a low ROE business.
Banks sell 23% of all annuities sold. At present, most of the insurance business that banks do is the sale of annuities. Here is a breakdown of insurance sales done by banks in the US (from LIMRA, as reported in the National Underwriter):
|Product||Percentage of Sales|
|Benefits and other commercial lines||16%|
|Individual Life and Health||4%|
Aside from annuity sales, the other major area of insurance activity for banks is the brokerage of employee benefits and commercial insurance. Banks have been aggressive buyers of local insurance brokers; one-third of all sales of local insurance brokerages since 1995 have been sold to banks.
There is logic to banks engaging in insurance brokerage. It deepens commercial relationships within a bank’s footprint, and can even lead to opportunities to expand the geographic scope of a bank as it buys insurance brokerages outside its footprint. Insurance brokerage relationships can lead to new banking clients, and vice-versa.
There are two banks among the top ten insurance brokers in the US: Wells Fargo (WFC) is fifth, and BB&T (BBT) is sixth, behind the big insurance brokerage specialists Marsh and McClennan (MMC), Aon (AON), Arthur J Gallagher (AJG), and Brown and Brown (BRO). There are cultural differences between banking and insurance brokerage. A large commitment to insurance brokerage by a bank implies that the brokerage arm will behave like the big insurance brokerage firms that they compete with. Banks with a small commitment to insurance brokerage tend resemble the small brokers that the bank acquired. And in general, insurance brokerage tends to be a more aggressive sales- and customer service-driven culture.
We are not yet at the end of the involvement of banks in the insurance business. Intelligent bankers will use insurance as yet another way to deepen the relationships that they have with commercial, and to a lesser extent, retail clients. In general, we do not expect many banks to take on underwriting risk.
Bringing It to the Present
All of this is still true today, and banks don’t know what to do with insurance, aside from a few of them selling annuities, like CDs, and being insurance brokers through their business banking relationships.
The last major bit of the Travelers acquisition was unwound as well, as MetLife (MET) bought the Life & Annuity business of Travelers for an attractive price.
One correction: in general, we now know that insurers do asset-liability management far better than the banks, and that the banks were considerably overlevered compared to the stable insurers.
I still think the best summary here is: banks can be good marketers of insurance. They are a logical distribution channel for many lines. But they don’t do a good job managing insurers.
Insurers may be better at managing their own pup banks, like Allstate and MetLife, but the length of the time of success is too short to be definitive. Be skeptical of large efforts to blend banking and insurance; it usually doesn’t work.
Disclosure: Author long ALL, SAFT, CB, PRE and RGA