Front Book, Back Book: A Framework for Understanding Financial Companies

Plus: Infinite Games and Wirecard, European Bank Consolidation, Chinese State Council

Issue #5 of Net Interest, my newsletter on financial sector themes, and I am overwhelmed with the response. Jamie Dimon of JPMorgan likes it, Nick Maggiulli of FinTwit likes it and scores of top investment managers like it. If you're reading this but you're not subscribed, you can join them by signing up here. You’ll receive Net Interest direct to your inbox every Friday.

Forwarded this? Subscribe here

Front Book, Back Book: A Framework for Understanding Financial Companies

Big banks and insurance companies are a thing of the past.

Not in the way that fintech founders think (that’s a topic for another note). Rather, their balance sheets reflect decisions that were taken long ago, sometimes a very long time ago. 

Warren Buffett has often spoken about the ‘long tail’ business that Berkshire Hathaway writes — policies generating claims that often take many years to resolve. In his 1985 shareholder letter he told how an advertisement placed in an insurance weekly had garnered US$50 million of premiums. But he warned: “Hold the applause: it’s all long-tail business and it will be at least five years before we know whether this marketing success was also an underwriting success.”

It’s the same at banks. The average life of a mortgage in the UK is seven years. A mortgage underwritten in between bouts of the Harlem Shake will still be on a bank’s balance sheet today. Even an investment bank trafficking in liquid securities can end up with assets lodged on its balance sheet. This time last year Deutsche Bank set up a special unit to ring-fence such assets that it wanted rid of — the average life of a portfolio of interest rate derivatives in there was eight years. As a benchmark for how long eight years is, it’s worth remembering that Deutsche Bank has had four CEOs in that time.

As banks and insurance companies get older, they accumulate a larger ‘back book’ of business. A large back book has advantages. A financial company doesn’t have to start all over again on 1 January each year. Before it even opens its doors it can bank on a stream of income coming in (although a stream of charge offs or claims could very well go the other way). Unlike other businesses a back book makes money during the night and over the weekend. It’s a melting ice cube of course, the asset won’t stick around forever, but as long as it was priced correctly, the back book can provide a stable stream of earnings to finance new business growth.

Problems arise however, when pricing on the back book is off. Berkshire Hathaway’s reinsurance business General Re suffered a big loss in 2001 because of this, in the aftermath of 9/11. Buffett explained, “it did not reserve correctly…and therefore severely miscalculated the cost of the product it was selling. Not knowing your costs will cause problems in any business. In long-tail reinsurance, where years of unawareness will promote and prolong severe underpricing, ignorance of true costs is dynamite.”

Just as no-one priced in the possibility of large-scale terrorism losses in 2001, no-one priced in the possibility of pandemic losses in 2020. Lloyd’s of London estimates that insurance industry losses will exceed US$100 billion this year. That’s around twice what 9/11 cost (US$40 billion or US$55 billion in today’s money). 

Again, it’s the same at banks. They are expected to incur significant loan losses over the next year or so from customers unable to pay back pre-pandemic loans. The Bank of England estimates that UK banks alone could suffer £80 billion of credit losses as a result of the pandemic. 

The possibility of incurring losses on back books makes them a very capital intensive enterprise for financial companies. [1] But they’re pretty cheap to maintain, so the operating expenditure associated with them is quite low. Front books are the opposite. It’s quite expensive to win new business, but not much incremental capital is required to do it. This means financial companies consist of two discrete businesses:

  • A back book, representing the past — high capital intensity, high operating margin

  • A front book, representing the future — low capital intensity, low operating margin

Financial services isn’t the only industry to have evolved this way. It’s just that the financial industry’s back book is much bigger. Consolidate the gross assets of S&P 500 companies and financials make up over half the total in spite of making up only an eighth of the constituents (and a tenth of the market cap, but that’s another story).

Other sectors have finessed the back book model. The genius of SaaS is to have created a back book with low capital intensity and high margin. But the origins can be traced back to financial services. This week Tren Griffin tweeted: “The microeconomics of a software as a service (SaaS) business evolved from cable and mobile. John Malone invented what is the most common business model for SaaS.” John Malone, in turn, borrowed heavily from real estate:

“The concept that cable television looked more like real estate than it did manufacturing was always obvious…to me, anyway. And I think the financial markets really didn’t have a model for cable, because the industry was a small, startup industry with no real following. Coming out of that period of the ’70s, the industry needed some model, some metric how the market could value us. We decided…to go on a cash flow metric very much like real estate.”

Cleaving apart the businesses

In spite of their high operating margins, high capital intensity can crush back book returns. McKinsey estimates that the provision of balance sheet and fulfillment functions that constitute their back book make banks a 6% return on equity (ROE). The front book meanwhile, consisting of origination, sales, distribution and other customer-facing activities, makes them a 22% ROE. 

Any gap of that magnitude is going to attract arbitrage and the way to effect it is to cleave the businesses apart. As Jim Barksdale said, “In business, there are two ways to make money. You can bundle, or you can unbundle.”

In financial services this really got going via the process of securitisation in the 1980’s, with the US mortgage market the epicentre. The people who originated mortgages no longer had to be the same as the ones who serviced them, and the people who serviced them no longer had to be the same as the ones who underwrote them. Michael Lewis wrote about it in his classic, Liar’s Poker: “thrifts became traders and traders thrifts.”

For the value chain to be cleaved as cleanly as it has been in the US a deep capital market has to be present to provide the financing role. Last week I looked at India as a case where the capital market is not sufficiently deep. Housing finance companies emerged to do the origination job but because they are reliant on bank loans for funding they are not truly independent. The resilience of the universal banking model in Europe owes a lot to the still relatively underdeveloped capital markets here. In spite of their promise after the creation of the Euro, capital markets never did achieve the same levels of penetration as in the US. 

In the insurance industry the reinsurance market plays the role of the capital market. Lemonade details in its S-1 prospectus how beginning 1 July it will cede 75% of its business to reinsurers. Like most fintechs it does not have the capital to cultivate a back book. Quicken Loans, which also announced an IPO last week, similarly sells off its origination, operating an entirely front end oriented mortgage business. 

Sometimes in trying to cleave apart the businesses, financial engineers can get a little too cute. The problem of course is that it removes skin in the game. If you’re not going to retain some interest in the long-term performance of the book, there is little incentive to underwrite it properly at the outset. The industry employs several devices to address this. These include retaining a share of the risk, like Lemonade does, and issuing reps and warranties, like Quicken Loans does.

A third, weaker, device is reputation. This is the one used by marketplace lenders. It is apparent that these have not been successful businesses. Funding Circle is back at the price it was five funding rounds ago in October 2013, Renaud Laplanche’s second act Upgrade is now worth more than his first, Lending Club, after a capital raise this week and UK marketplace lender RateSetter has put itself up for sale. The tail associated with personal lending may not be as long as catastrophe insurance but it’s longer than merchant returns, making this an inherently different business from others where bits are moved around. And reputation is too vague a device to align incentives.

Looked at the other way, bundling the back book with the front book is a low-friction way to address the agency problem. 

There’s also a diversification benefit in keeping the two books together. It’s a basic economic principle that price serves as a rationing mechanism. When capital is abundant, pricing goes down. At times like those, a well-underwritten back book can sustain the business until the cycle turns. If the entire business is built around flow it’s harder to withstand competitive forces. This is what Chuck Prince meant when he said “As long as the music is playing, you've got to get up and dance.” But when capital is depleted, like it is after large losses and pricing goes up — that’s really the time to dance.

Right now insurance pricing is doing exactly that. This week US wholesale broker Amwins published in its State of the Market report that renewals on property insurance are increasing at double digit rates versus last year, with insurers “taking ongoing and additional steps to improve profitability”. Where they can, insurers are raising capital to exploit this dynamic. Renaissance Re, which raised US$900 million this month, wrote in its prospectus: “We believe that the COVID-19 pandemic is accelerating the recent rate increases we have seen in many of the lines of business that we write. In addition, the market has been impacted by a tightened supply of capital.”

At the other end of the spectrum, a pocket of the financial sector where capacity has increased lately is aircraft leasing. These companies own over 40% of the global fleet of passenger aircraft, which they lease back to airlines. The largest, Aercap, owns 934 planes and has another 342 on order. The outlook for air travel is highly uncertain, but as long as there are too many planes in the skies (or rather on the ground) lessor pricing power diminishes.

This letter doesn’t make investment recommendations. But it does make observations. The way to make money investing in a front book oriented business is to follow the pricing higher. This works especially well in a consolidated industry. The way to make money investing in a back book is to value it higher than the market. Current uncertainty makes the latter harder than the former. In some ways it can be easier to predict the future than the past.


[1] In all the controversy surrounding how much capital financial companies need, people sometimes forget what financial company capital is for. It’s not there for expected (priced) losses — that’s what reserves are for. It’s there for unexpected losses.

Forwarded this? Subscribe here

More Net Interest

Infinite Games and Wirecard

“There are at least two kinds of games. One could be called finite, the other infinite. A finite game is played for the purpose of winning, an infinite game for the purpose of continuing the play.” — James Carse

We are all taught that infinite games are a better model for business and for life. But the incentives engendered in the model aren’t always appropriate. An accountancy firm wants to do a proper audit but it also wants to retain its clients’ business. A sell-side analyst wants to do a proper evaluation but he also wants to maintain access to management. A regulator wants to ferret out fraud but they also want to protect their reputation. In each case the infinite game undermines the task at hand.

One group of market participants who don’t play infinite games are short-sellers. Critics accuse them of only wanting to make a fast buck. But that’s precisely why it works.

European Bank Consolidation

The ECB held its first post-Covid cheap funding auction this week and a staggering €1.3 trillion was taken up. Some of this (~€760 billion) was used to refinance maturing facilities but ~€550 billion was fresh funding, channeled from the ECB to banks’ balance sheets at a cost of -1%. Although big, the €1.3 trillion wasn’t the most surprising number in the release. That number was 742 — the number of banks that bid for funds. Who knew there were so many banks in the Eurozone?

Intesa Sanpaolo threatened to kickstart a merger wave in February when it announced a deal to buy BPER Banca in Italy. The merger is currently going through the antitrust process. Coming out of Covid, when banks realise they don’t need such a large physical presence, further consolidation is likely. What’s more, if equity valuations don’t recover, banks may be able to use negative goodwill to cover restructuring charges.

Chinese State Council

In Europe they use roundabout funding structures to get banks to lend (see above). In China they’re a lot more direct. The State Council this week announced: “China to encourage financial institutions to make interest concessions and expedite fee cuts to boost real economy.” The State Council explicitly targets a RMB 1.5 trillion transfer of value from the banking sector to the corporate sector. Perhaps because they’ve done this before they know that for this to work their banks need to have built up a store of profits to transfer over. Unlike in Europe, Chinese bank profitability has been quite strong over recent years. European policymakers are converging on the same place, but they’re a lot more procyclical.