Impac Mortgage: Where the money comes from
When I first bought Impac Mortgage (back at the beginning of August) it was on the basis of GAAP earnings (which were 50 cents per share in the second quarter), and revenue growth from the mortgage origination business. Soon after, when I looked more closely into earnings, I determined that much of what was reported in GAAP was obscured by mark to market adjustments and a legacy business that is no longer operating. Fortunately if I ignored these effects, the resulting picture was even better than the one painted by GAAP.
So I left it alone and went on to other things.
To digress for a minute, this is my process. Once I feel like I have a clear answer on a stock, I don’t look too much further into the details. When I look at a stock I look hard, and I usually come up with a fairly accurate picture, but after I feel able to draw a conclusion, I don’t spend a lot more time quibbling over the details.
I don’t have time. I have time to look into maybe 2 stocks per week. If I spent week after week evaluating a single security, it is simply inefficient.
Does this lead to mistakes? Absolutely. Sometimes I miss a key aspect that changes the equation. But to mitigate mistakes I have learned to reevaluate when the market tells me I am wrong, and to act quickly when it turns out I am. And actually, this has been one advantage of starting this blog. There have been a couple of cases where readers have pointed out something that I have missed. And I’ve saved money as a result.
Given the amount of time I have to allocate to investing, this remains, in my opinion, my most efficient process. Study the business, figure out what the key drivers are and where problems are most likely to arise, evaluate those drivers and problems, make a decision and move on to the next one. Take another look if things start to go amiss.
With respect to Impac, as the stock moved up from $2.50 to $10, I wasn’t that concerned with getting a better grasp on the specifics of earnings. My initial analysis showed me the drivers, and they led me to conclude that the stock wasn’t even close to reflecting those drivers, and that was enough for me.
But now, with Impac hovering between $10 and $11, further analysis is warranted. My intent below is to understand how each of the businesses that Impac operates generates earnings, and to compare the earnings generation capacity to GAAP, hopefully eliminating some of the confusion introduced by GAAP.
How Impac generates revenue
Impac Mortgage generates revenues from 3 different sources:
- Cash spun off from the residual trusts
- Mortgage origination fees and gains
- Real estate service fees
I have already written in some detail about the cash generation of the residual trusts, so I am not going to spend any time on that here. Instead I will focus on the gains from mortgage originations and from real estate service fees.
The Origination Business
The mortgage origination business is the one I am most interested in. The company originates mortgages through wholesale, correspondent and retail channels. The mortgages originated at in compliance with the agencies, and are generally either sold to Fannie and Freddie, or packaged into securities that satisfy the requirements of Ginnie Mae.
Let’s talk about each of the origination channels:
Retail lending is when the company deals directly with the borrower. The retail business is more often the highest profit business (because there is no middle man taking a cut) but it also requires more up front capital and fixed costs because you have to develop the infrastructure, the marketing, the people and the relationships required to interact with the borrower directly. Higher start-up costs were at least partially responsible for the losses incurred in previous quarters.
In the wholesale lending business, Impac acts as the middle man between the originators (brokers) and the buyer (one of the GSE’s). Impac gives the broker a somewhat lower rate than they would a customer in the retail channel. The broker takes their cut, and Impac provides the funding for the loan (through its warehouse lending arrangement), review the loan for compliance with the GSE criteria, and if it is compliant, send it off for securitization. The result is somewhat lower margins, but also somewhat less work and infrastructure.
The counterparties in Impac’s correspondent lending are community and regional banks, credit unions, and some approved lenders. The correspondent lenders aren’t using Impac’s funding lines and thus they alone are making the decisions on whether the loan should be funded and at what rate. Once the loan has been funded, they sell the loan to Impac and Impac does whatever additional work needs to be done to get it ready for the GSE’s (such as stripping out the servicing piece to either keep or sell separately) and probably more importantly, takes the loan off of the hands of the lender, freeing them up to make additional loans. Because the majority of the origination process is being done by the lender, this business is lower margin to Impac than Retail and Wholesale.
As I have already pointed out in a previous post, each of the three arms of the business are growing.
I did some work to look at the margins Impac is making on the business. While the data available is sparse, as Impac does not provide a detailed breakdown of their gain on sale, I was still able to pull together the following.
I also compared Impac’s origination margins with PHH Corp and Nationstar. Below are the numbers for the second quarter.
Impac seems to account for their gain on sale a little differently than PHH or Nationstar. Impac breaks out the realized gains on derivative instruments as a separate item, whereas I believe that Nationstar and PHH are simply adding that directly to gain on sale (which makes sense since the purpose of those derivatives is to protect the value of the margins on the interest rate lock commitments from the time they are made to the time they are sold, though I didn’t find specific reference to this). Therefore, to do a apples to apples comparison, the realized gain on derivatives should be added to gain on sale for Impac. It illustrates that Impac is operating a lower margin business than PHH and Nationstar. This isn’t surprising since Impac is deriving a significant amount of their business from correspondent lending, where the margins are small.
A second difference is that both PHH and Nationstar capitalize servicing rights, and that capitalized gain is added to gain on sale. I believe that Impac accounts for servicing under the line item other non-interest income, and thus don’t show up as an item within the gain on sale. If I’m right about this, in the second quarter the servicing segment saw a small loss ($632,000) but most of that loss is likely due to changes in fair value, offset by a capitalized gain on new servicing rights. With the information provided I don’t think I can estimate what value Impac is capitalizing servicing at. At any rate, Impac’s comparative margin on gain on sale is going to be between 60-100 basis points lower than Nationstar and PHH because of the differing accounting treatment of servicing.
After adjusting for the differences in accounting, gain on sale and costs seem to be pretty much in line with what I would expect. Going forward, I would expect that both the gain on sale margin and expenses will increase as the retail segment takes on a bigger percentage of volume.
I’m not really sure how much in the way of potential for margin expansion there is. Impac does have lower overall margins than PHH and Nationstar, which are both much bigger players in the arena, and so there may be some improvement in margins as Impac increases in size. But I know from my experience with Nationstar that margin expansion is not a big driver in this business. Impac is really a volume play; the origination segment is going to see growing revenues as it expands it footprint and takes advantage of the recovering housing market. Impac is on their way towards more than quadrupling origination volumes over 2011. While that has to slow going forward, its still likely to show significant growth in 2013 and beyond.
Real Estate Services Business
While the origination business is the growth business, the real estate services business also generates a significant amount of operating profits. Unfortunately, it wasn’t until the second quarter that Impac began to break out real estate services from origination, so its tough to get a strong sense of its historical profitability. The only data available is revenue.
The company defines its real estate services as providing the following functions:
- REO surveillance and disposition services to portfolio managers and servicers to assist them with improving portfolio performance by maximizing liquidation proceeds from managing foreclosed real estate assets, and short sale (where a lender agrees to take less than the balance owed from the borrower) services on pre-foreclosure properties for servicers, investors and institutions with distressed and delinquent residential and multifamily mortgage portfolios including real estate brokerage services.
- Default surveillance and loss recovery services for residential and multifamily mortgage portfolios for servicers and investors to assist them with overall portfolio performance and maximizing cash recovery.
- Loan modification solutions to individual borrowers by interacting with loan servicers on behalf of the borrowers to assist them in lowering the monthly mortgage payments to an affordable level allowing them to remain in their homes. The Company receives fees paid by the borrower for these services.
- Monitoring, reconciling and reporting services for residential and multifamily mortgage portfolios for investors and servicers.
Impac primarily provides these services for the long-term mortgage portfolios it originated and spun out into trusts. Therefore there has to be an expectation that the revenues from this segment will decline going forward as those portfolios wind down if Impac cannot find third party business. The company said this about the business prospects in the recent 10-Q:
Although the Company seeks to expand its portfolio loss mitigation and real estate services to more third parties in the marketplace, the revenues from these business activities have historically been generated from the Company’s long-term mortgage portfolio. Furthermore, as the distressed mortgage and real estate markets remain unstable and uncertain due to the number of foreclosure properties that need to be sold, there remains uncertainty about the ongoing need and delivery of these services in the future
You can see the list of foreclosed properties that Impac is selling here.
What occurs to me is that the above list of services is not that far off what a mortgage servicer provides. While I am only speculating, I have to wonder whether there is the potential to transform this business into a servicing business. Right now Impac uses a sub-servicer to perform the servicing tasks.
Discontinued operations consist of the Alt-A mortgage origination business that the company was in prior to 2008. While they have not been originating Alt-A mortgages for over 4 years, they continue to receive repurchase requests from Fannie Mae for mortgages that breached the representations and warranties agreed to when they were originated.
In the last quarter the company paid $1 million to settle breaches of representations and warranties, and the reserve held for future payment increased another $1.8 million. This means that the company added about $2.8 million to reserves for future repurchase demands. This led to a significant loss of $3.1 million for Discontinued Operations. In the first quarter a repurchase reserve of $500,000 was added.
Impac doesn’t provide a lot of information on what to expect from discontinued operations going forward so its difficult to know the future losses are going to be. By way of analogy, PHH does provide a lot of information with respect to how their reps and warranties reserves will change going forward. PHH has taken a big hit to earnings from increases to their repurchase reserves over the last two quarters. On the second quarter conference call the company said they expected repurchases to remain elevated until the end of 2013. The company also provided an estimate of possible repurchase reserves; these are loans that are not yet reserved for but that may require reserving in future quarters. Possible reserves have been declining but they are still high historically; if one were to assume that all the possible reserves became reserves over the next 6 quarters (so until the end of 2013), PHH would be reserving at a level only slightly below where it reserved in the second quarter.
Such a comparison would imply that Impac will continue to add to its repurchase reserves over the coming quarters, but that the additions will probably begin to taper off shortly. Still, it’s really difficult to know how the repurchase reserve is going to change going forward. In particular, was the first quarter representative of what to expect, or was the second quarter? The answer to that question is going to have a big impact on GAAP.
An optimistic view would assume that the company decided to take a large one-time hit to earnings in the second quarter in order to get its repurchase reserve in line with expected repurchase. A pessimistic view would wonder if the the second quarter is merely the first in a series of large writedowns. I don’t know the answer to this; its simply a risk that will have to be monitored.
All of the work that I did above and in my previous post on the trusts was tedious, but it was necessary in order for me to understand the business. And I needed to understand the business before I came up with an estimate of its overall profitability.
In the table below I have looked back over the last 6 quarters of earnings, ignoring the effect of the mark to market changes on the non-recourse trusts and the discontinued operations. The result is the income of the operating businesses.
I am interested in operating earnings. Therefore I am ignoring the effects of the discontinued operations and of the trusts, including the cash residuals that Impac receives from the trusts. I have included the cash residuals per share on another line item below. Technically, these are a part of the actual earnings Impac receives from the trusts they hold, and they should be included as such (instead of the mark to market adjustments that get included as per GAAP). However because they are uncertain going forward as these trusts wind down, I’ve decided to break them out separately.
Operating earnings were around 80 cents per share in the second quarter.
The conclusion that I draw from all of this is simply this: I don’t think that the current price of Impac Mortgage reflects its earnings potential. While the short term GAAP results will continue to be impacted by changes to the repurchase reserves, and obscured by the mark to market changes of the non-recourse trusts, the basic origination and real estate services businesses are beginning to grind out some pretty impressive operating earnings. The 80 cent per share number that I estimate should be sustainable and even grow going forward as the origination business grows. On top of the that, the company continues to generate cash of over $1 per share from the residuals trusts. Worth noting is that Impac has pretty significant operating losses that they are carrying forward, so I don’t expect that taxes will be significant for quite some time.
In my opinion, the biggest short term risk is that the third quarter GAAP number disappoints. This could easily happen as a result of further charges to the repurchase reserve or mark to market changes to the trusts that overwhelm the profitability of the operating business. Given the run-up in the stock, a low GAAP number could cause some of the short term holders that haven’t looked closely at the businesses to run for the exits.
However, it equally possible that the third quarter surprises to the upside. On top of growth from the origination segment (particularly to the retail business that is expected to show growth in Q3), there is the possibility that the change to the repurchase reserve was a one time event in Q2 and will not repeated. The mark to market of the trusts is also a big wild card; there could be GAAP upside from a re-evaluation of trust assets predicated on rising home prices. If all of these effects occur, the GAAP number could be outstanding.
In either event I don’t plan to sell any shares. What I am going to focus on are the operating businesses. I want to see growth from origination, and a gradual decline from real estate services. As long as both of these occurs, I plan to hold and maybe add to my position going forward.