I am a big fan of title insurers. As I discussed in my earlier write-up on Investors Title Company (NASDAQ: ITIC), title insurers receive a single premium up front that insures the policyholder in the event that a property’s title is not clear at the time of purchase/mortgage. So, imagine you buy a house and put a mortgage on it, only to find out later that the fellow that sold you the house didn’t actually have the right to do so – he owned the house with his estranged wife who is now knocking at your door demanding cash. If you had bought title insurance, you are covered. If not, you are out of luck (though, the bank that put a mortgage on the house forced you to pay for title insurance, as part of their fees).
So, you might be asking: Isn’t this an inferior insurance model? Who wants to receive a single payment up front for a liability that may not be uncovered for decades? I like the model for a few reasons. First, the incidence of paying out on these policies is relatively well understood. The vast majority of claims are made within the first six years of the policy. Yes, there is a long-tail of claims, but unlike something like Asbestos, or Japanese earthquakes, the long-tail probability has a low associated cost. This is the distinguishing factor that I think makes title insurance attractive. There will be a small number of claims made many, many years after the policy is taken out, but these claims will be relatively small and in the meantime, the insurer gets to invest that money for its own use.
First, how do I know these claims will be relatively small? The bulk of title insurance policies are taken out on behalf of banks who want to be sure the mortgage they are putting on the property will be paid off in the event that the property does not have a clear title. Over time, the property owner pays down this mortgage, which reduces the amount of the potential liability for the title insurer. In the event of a refinancing, or property sale, the liability is extinguished. So, when interest rates are low and homeowners refinance in relatively greater numbers, title insurers (in the aggregate) gain on two accounts: first, their liabilities are extinguished upon sale, and second the new mortgage needs more title insurance, which increases revenues. I’ve written “in the aggregate” here because the new policy may be taken out from a different firm, but in fact the title insurance industry appears to be dominated geographically by one or two firms. For example, as I mentioned in the ITIC write-up, that company essentially controls the North Carolina title insurance market.
I should note two other things: First, title insurers set their reserves based on the undiscounted expected liabilities.
Ok, enough with the primer on title insurance. You’ve seen why I like the industry, but what about particular companies? In ITIC’s case, investors at the time would essentially have been getting the business for free, as the company could have paid out around 2/3 of its investment portfolio (which approximated the company’s market cap at the time) without jeopardizing its ability to continue to pay claims. This was the result of fifteen years of over-provisioning relative to claims. Since I did that write-up on 3/22, ITIC is up around 28% (as of 6/8/11). Given the run-up, ITIC no longer presents as compelling an opportunity, but I think Stewart Information Services Corporation (NYSE: STC) is in a similar situation as ITIC two months ago.
STC is a title insurer that has been in the title business since 1893 with operations in 38 US states, Canada, UK, Mexico, Central America and Australia. The company’s current claims reserve stands at nearly $500 million, which is more than three times its three year average actual claims net of recoveries, which I argue are inflated figures given the leasing shenanigans we’ve seen (the three year average is nearly double the prior five year average). We will see a mean reversion in the claims, and in response to the increased claims, the company increased its reserves accordingly over the last few years (I think this is a good sign of conservative management, as they could have simply run down their reserves and maintained earnings artificially high). In fact, looking back over the last seventeen years of available data in the SEC EDGAR database, we see that STC has made provisions in excess of actual claims for sixteen of those years. These provisions have averaged more than 40% more than actual claims over the period.
As I noted in this post, a perhaps better way of looking at the reserves is to consider STC’s reserves relative to last year’s claims and trailing three year claims. Here’s the graph:
So, from this perspective, it appears that the company is becoming somewhat less conservative, especially in relation to the late 90s. However, I think the reserves are still quite adequate.
So is it cheap?
The company’s book value of equity is $433 million ($445m less $13m for noncontrolling interests), or 123% more than the company’s current market cap of $194 million. I believe BV of Equity is a fair approximation of current market value of assets because the company’s assets are highly liquid. Additionally, I think that the liabilities are somewhat overstated (since the company has been overprovisioning for nearly two decades).
Should the company be trading at such a discount to NAV? Some discount may be justified since the company has been losing money for a few years. But I argue that those losses are due largely to a rapid increase in provisioning (the title loss expenses increased from an average of 4.9% from 1995 to 2006 to an average of 10.1% from 2007 – 2010). I expect these losses will mean revert given the changes they’ve made to their agency network (more on this below). Additionally, the company has made substantial improvements to its cost structure, reducing headcount, improving technology and transitioning to Regional Production Centers to reduce expenses. This is an ongoing process, but we have already seen significant improvements (see graphs below) which will lead to improved profitability.
I argue that in the near term the company should benefit from several things:
First, loss provisioning will return to normal levels, which will reduce that expense by ~$80 million. Also, I should note that the company should take a gain this next quarter to reverse the “reserve strengthening” charges they took in relation to a lawsuit it was engaged in with Citigroup. That lawsuit was resolved this quarter in favour of STC, so that money should flow back as a gain.
Second, the company’s real estate information division has been growing like crazy over the last few years and the trend appears to be continuing this year, with revenues this last quarter amounting to nearly 40% of the division’s full 2010 revenues. Though this division is a small part of the company’s overall business, it doesn’t take much to swing EPS given that the company has just 19.4 million shares outstanding.
Third, the company has been cutting its employee and other operating costs quite successfully in both divisions, as the following charts show. The company is continuing to transition to Regional Production Centers, which will lead to further improvements in employee expenses.
Fourth, STC has been renegotiating its remittance contracts with the agencies it deals with. This increased remittance rate will flow directly to STC’s bottom line. So far, they have successfully increased remittances in 38 states.
Fifth, STC has been increasing its premiums without realizing a concomitant decline in revenues (more evidence of geographic control in the industry). They have raised premiums in 27 states so far.
Sixth, the company has improved its agency network. A weak agency network is what led to the massive increase in actual claims over the last few years. Losses on known claims from independent agencies has been reduced from 9.2% a few years ago to 2.9% today, and more than 70% of those claims were from agencies STC has recently stopped working with).
For these reasons, I am a fan of STC. If you’ve looked at STC, I’d love to hear your thoughts!
Author Disclosure: Long STC, ITIC