Radian's (RDN) CEO's performance at the Barclay's Investor Conference on 11th September 2012 has compelled me to write this article. The CEO is misleading investors - not illegally, but he is misleading them. On the call today, in response to questions, he said that Radian is one of the best reserved companies in the industry (in terms of reserve per delinquency) and that their biggest investors ask them why they don't bring down reserves. While the CEO is not lying (it is true they have one of the highest reserves per deliquencies in the business), their reserve per delinquency is artificially inflated by various tricks I describe in the following excerpt from a report I wrote on MGIC Investment Corporation (MTG) - Radian's competitor). He speaks in a way that's unfair to investors who don't know better.
Furthermore, Radian's mortgage insurance writing company liquidity position is dangerous, similar to PMI Group's (PMI) before it was seized. While I do not expect Radian to be seized or go to zero, I expect there is a lot of money to be made by going long MTG and short RDN, even if MTG is forced into run-off by the Freddie dispute. However, it is still possible that RDN's holding company gets into trouble, even if the regulator doesn't raise any red flags about their operating subsidiary liquidity. I doubt the regulator will allow them to upstream cash to take care of their debt obligations, even if the mortgage market improves. They will probably need all their cash to pay claims, maybe even more than they have, and so there will be no cash to send to the holding company.
So I believe that the CEO wants to push the stock price up to a level to raise capital from unwitting investors - investors that clearly don't understand the difference between the mortgage insurance stocks, as evidenced by the excerpt below. Who are these "big investors" asking management why they are over-reserved? They clearly don't understand the business they are invested in.
Note that since writing this report, Radian's monthly data showed paid claims surging, whereas MTG's stayed level.
The rest of this article is an excerpt from my MTG investor report, published 6th August 2012. Originally I wrote a section on Radian to show that the market does not understand how to compare these mortgage insurance companies.
But the most convincing evidence that the market just doesn't "get" these stocks is its current appraisal of Radian Group. RDN should not command such a high premium to MTG and I am about to show you why. Hopefully this will convince you that the market doesn't understand these names.
Radian was one of the earliest mortgage insurers to suffer. As soon as MTG saw the crisis coming, it raised capital. Radian didn't. As a result, the market dumped Radian shares, believing it was only a matter of time before bankruptcy.
But Radian had a trick up its sleeve - its holding company owned another business, a bond insurer. Bond insurers write protection on municipal debt and structured products. Unlike other bond insurers, management had chosen not to insure mortgage bonds at Radian Asset Assurance - RAA - because it didn't want the fates of its two businesses to be intertwined. RAA had a lot of surplus capital and so Radian Group (the holdco) gave the company to its mortgage insurance subsidiary, Radian Guaranty - RG. The regulator allowed RG to use the capital of RAA when calculating capital ratios. Radian, shut out from the stock market, no longer needed access. The stock surged.
But the problem is this: while financial statements make it look like Radian Guaranty has a lot of cash and investments, a significant portion is locked up in RAA. Radian is only allowed - by statutory law - to extract a small sum of cash from RAA each year. While on a risk to capital basis RG looks very strong, on a liquidity basis, it is very weak - almost as weak as PMI.
And since regulators care less about risk to capital than they do about liquidity, RG is on shaky ground.
Radian's capital waiver, for instance, requires that RG keep at least $0.7 million in cash and investments. RG's current position? Its statutory filings (not its SEC filings where it counts the illiquid financial guaranty stake under "cash and investments" - again not a lie but definitely misleading) show that RG has $2.3 billion in cash and investments. MTG has $6 billion (including trust assets). Since MTG has 50% more risk than RG, Radian would need $4 billion in cash to be on equal footing. $2.3 is a far cry from $4.
Radian - it seems - is aware of the problem and has borrowed a page out of PMI Group's playbook (a competitor that was seized by their regulator last year). Once PMI realized it had a cash problem two things happened. First, it started increasing the portion of its business that was single premium, where the borrower pays a single lump sum at the beginning of the mortgage. PMI did not announce its new business strategy - it was only revealed by the balance sheet - through an increase in unearned premiums. Second, it started denying claims at an accelerated pace, making claims look lower than they actually were. These actions showed that PMI was scared.
Fast forward to today and Radian is playing the same game. Denials are through the roof, far higher than the other publicly traded mortgage insurers, who have no need to treat their customers that badly. Most significantly, single premium business is sky rocketing. Radian is undercutting the competition with horrible prices because it is desperate for cash. Their prices are equivalent to asking a borrower to pay 2.5-3.5 years (depending on FICO and LTV) of monthly premiums. The problem: with interest rates low, and house prices rising only gradually, these mortgages are going to be on the books for a very long time - probably more than five years. To make matters worse, interest rates are so low that they won't generate much investment income with the cash.
Radian is only offering these prices because it needs cash - it needs to convince the regulator that it will have enough to pay policyholders on time. The regulator does not want to give shares in Radian's bond insurer, RAA, as a substitute for claims.
But who knows? Maybe Radian's tactic will work. They are generating about $720mn in premiums per year and $680 million in pre-claim cash flow. If claims rise to $1.6 billion - as they should once the denial trick fades (they should actually temporarily rise higher as they "catch up" on delayed claims) - they will burn $900 million per year. That gives RG 2.5 years before it runs out of cash and 1.8 years before they breach the minimum cash and investment requirement. Sound similar to PMI? As soon as claims start to rise the regulator will have to take notice. Radian longs should be very careful.
Even ignoring the fact their bond insurer insures a ton of risk that could lead to capital losses at any moment, Radian's problems don't finish there. The market likes to assess the quality of these companies' loss reserves (the reserves they put aside for claims on current delinquencies) by looking at the ratio of loss reserves to delinquencies. MTG, for example, has $3.9 billion in reserve and 153,000 delinquencies. Their "reserve per delinquency" is $25,500. Radian, on the other hand, has $2.8 billion in reserve on 98,000 delinquencies. Their "reserve per delinquency" is $28,500.
So Radian is in a much better position than MTG, right? Not only does it have a stronger capital position, but it has higher quality reserves. MTG is going to need to catch up and will further deplete its already weak capital position by doing so.
No. Radian's loss reserves are a smoke screen. There are a number of factors at work that mean MTG's reserves are higher quality than Radian's.
The first fact to consider is that Radian has been slowing down claims. Because of this, "pending claims" are growing as a proportion of Radian's reserves (see the quarterly slides). "Pending claims" almost certainly end up in a claim, which means Radian should have a higher loss reserve than MTG. We can adjust Radian's reserve for comparison with MTG. Only 7.8% of MTG's delinquent loans are in the pending claim category (see MTG quarterly release) vs. 17% for Radian. Radian should have already paid about 9,000 claims, which would result in a reserve per delinquency of $26,500 (and almost $500 million less cash, which would drain its mortgage insurance co to $1.7 million of cash and investments, which is dangerously low).
The second fact to consider is that Radian is boosting reserves to account for denial reversals. Radian is assuming that some of its denials are going to return as claims, and is reserving appropriately. Since these loans are not delinquent yet, Radian must use a special reserve account, "IBNR". On first reflection this might sound like a point to Radian - they're boosting reserves. But it's not - they are artificially inflating their reserve per delinquency by boosting the numerator (the loss reserve) in absence of a rise in the denominator (number of delinquencies). The future reversal of denials is not currently counted as delinquent loans. Radian has $220 million IBNR and MTG has $100 million (see recent 10-Qs). Adjusted for difference in size, Radian should have only $65 million. Thus Radian's loss reserve should be $155 million lower, which reduces reserve per delinquency from $26,500 to $25,000.
The fun does not stop there. Radian's book of business is also more risky. Radian has a higher proportion of loans in the worst states: 11.8% of Radian's risk in force is in California vs. 7.3% at MTG. A riskier book should result in a higher average claim payment (riskier loans tend to be higher balance), which I believe Radian has started to cover up. During the first 6 months of 2011, Radian's average claim was $55,000. MTG's was $48,000. During the first 6 months of 2012, Radian's average claim was $50,000. MTG's was $49,500. Why has this changed? Radian is probably denying higher balance loans to make its average claim look smaller. Think about it: if half of your claims are $25,000 and half $75,000, the best way to reduce your average claim is by denying the $75,000 loans. Radian's denial tactics are not going to last much longer; eventually the average claim should move back up to $55,000.
Since Radian's average claim is higher, their reserve needs to be adjusted appropriately. For every delinquency that goes to claim, Radian will pay 10% more than MTG. In other words, for Radian and MTG reserve quality to be the same, Radian's reserve per delinquency must be 10% higher. Thus Radian's true reserve per delinquency is actually $22,500 (cut $25,000 by 10%) vs. the $28,500 that it reports.
(Radian, if it wanted to clear its name in this regard, could release the dollar value of risk in force in default, as MTG does in its 10-Q. We could then compare the average balance of loans in default. But until they do it, it is prudent to assume they have something to hide.)
Finally, it is likely that the reversal in denials (clients rebutting) will be more significant than Radian thinks (more than the $155 million excess IBNR). Radian believes that only 50% will reverse, but why aren't other companies following the same tactic? Radian's Master Policy (their mortgage insurance contract) gives servicers only 3 months to rebut a denial. They then have 9 months before they can resubmit the claim. Other mortgage insurers have a longer rebuttal period, giving customers enough time to fix the defect. Expect claims to rise once the 12 month period (3 month rebuttal plus 9 month pause) starts to end***. These phantom delinquencies, which would further reduce their "MTG equivalent" reserve per delinquency, are a further scratch against Radian's name.
Will Radian's problems result in seizure (receivership) or an order demanding a 50% reduction in claims payments? While Radian is almost certainly suffering from a liquidity problem rather than insufficient claims paying resources, it might be 10 years before they can extract sufficient capital from the bond insurer. Policyholders are likely to end up with shares in RAA, waiting for an uncertain payout. Considering the extraordinary circumstances, the regulator will probably want to take control, and squeeze as much blood out of RG as possible, unless Radian can raise more capital.
Radian is in a worse position than MTG: its reserves are lower quality, it has the unknown of the bond insurance risk, and, most importantly, its mortgage insurance company has a liquidity position that envies PMI, which might result in a seizure, barring any miraculous recovery in the economy. Most of all, there is no reason the Radian premium should exist.
Note: Radian started increasing denials in 2Q11, as shown by these slides. Thus it makes sense that claims have only just started to surge this month, just over 12 months after they first ramping up denials (it takes a few months for new claim to be processed so lag is more than 12 months). Expect claims to continue to rise and cash burn to increase over the next few months.