|
Mike Carnes , SVP
Capital Markets Group
|
Rob Lee, SVP
Capital Markets Group
|
Alternative Servicing Fee Proposal
Wednesday, April 13th, 2011
A recent Mortgage Banking Servicing Summit in Washington, DC explored the issues that have shaken the mortgage industry and challenged the ways in which mortgages are repaid and serviced. The conference was attended by industry leaders, economists, academics, consumer groups and policy makers to serve as a forum for discussion on the housing crisis and its impact on the industry. At the forefront of the meeting were engaging discussions on mortgage servicing rights (MSRs) and the factors that contributed to servicer portfolio performance through the current crisis. Questions and commentary included topics on what has changed in the servicing markets, what new servicing fee structures are being proposed, and what impacts the latest Basel rulings will have on MSRs.
Recent efforts by the Federal Housing Finance Agency (FHFA) include the launching of a new initiative to consider alternatives for a new mortgage servicing compensation structure. MIAC would like to offer observations on the impact of a change to the servicing fee structure. The current servicing compensation model has room for improvement when it comes to delinquent mortgage servicing rights. Changing the servicing fee from its current form may have implications in terms of asset values, modeling risk, prepayment performance, hedging strategies, and capital and risk structures to name a few. To begin, we explore the major proposals on service fees. We can isolate the service fee structure options into four major categories.
- Existing Service Fee Structure: Leaving the current servicing fee (SFee) structure as is. Generic Service Fee (SFee) rate of 25 basis points on Fixed Rate product, 37.5 Basis points on Arms, 44 basis points on GNMA I Service fees;
- Lowering to a Minimum Service Fee: This would still a part of the balance sheet asset, but kept at a minimum of say 12.5 basis points;
- Structuring the servicing as 1.00% or some other small percentage of unguaranteed principal and interest on the serviced loan; or
- Creating a Fee for Service structure, where the servicer is paid a set amount for servicing performing loans and a set amount for servicing delinquent loans.
In considering the existing service fee structure, there has been debate over whether the current service fee levels are too high in comparison to the compensation for the services delivered. One argument is that due to ‘economies’ of scale, the servicers are generating too much service fee income premiums given the administrative work to service loans. Another argument is that incentives are not properly aligned and should be redesigned to reflect the reduced cost on current loans and the increased cost on delinquent loans. However, in discussions with many servicers, servicing costs have increased as the financial landscape over the past few years has continued to evolve. Increased regulatory compliance costs, delinquency advance costs, repurchase risk exposure, system modifications and increased time spent servicing the portfolio have all resulted in stress on the profit margins of many servicers. Historically, Servicing Fee revenue steams steadily increased as average loan balances on newly originated loans were on the rise. The additional revenue helped to offset increasing costs, but as loan balances began to decline, so did the profit margins
Many servicers’ business models were built on automated, turnkey processes to create economies of scale. However, once servicing portfolio performance started to deteriorate, associated expenses increased, while the income side of the cash flow servicing rights equation stayed constant or even lessened. In cases where delinquencies increased, servicing advances ballooned to advance the principal and interest on the delinquent loans. Performing Loan Servicing Fee Revenue decreased as more loans became highly delinquent. In general, on GSE loans, less late fee income was generated on the severely delinquent loans, float income decreased as market rates declined, and servicing income, which represents approximately 70% to 85% of the revenue side of the cash flow were reduced. As the chart below illustrates, service fee revenue in general declines over time while expenses increase. Increasing delinquencies and lower quality credit product caused by better quality product leaving the portfolio are some of the reasons for this occurrence.
In terms of the GSE servicer, many of the cash flow revenues have diminishing returns, meaning higher initial cash flows (revenues) with low initial costs (expenses) in the beginning life of the mortgage servicing rights portfolio. As the portfolio ages and delinquencies increase, the servicer collects less service fee revenue while his cost structure (expenses) increases to handle delinquent loans. Often, the servicing fee revenue streams in addition to the revenue from late fee income, escrow float earnings and cross selling opportunities that added additional ancillary income may not be enough to offset the expenses at a certain delinquency threshold.
As the chart below demonstrates, Service fees in general are front loaded while expenses are back-loaded, increasing as the portfolio becomes more delinquent.
The above graphs represent the impact of a portfolio becoming highly delinquent and reaching 40% delinquency levels by year five. In this stressed example, we used a 12.5 basis point service fee, a 10% CPR, and a $100 servicing cost. The graph below illustrates the same cost, CPR, and 12.5 basis point service fee with the exception of a much better delinquency profile -- one that represents delinquencies peaking at 5% and more reflective of historical delinquencies pre-crisis.
MIAC explores this topic in our article with examples using various cost structures, a lower servicing cost assumption for the large mortgage servicers and a higher servicing cost for the smaller servicers.
How would the participants or MSR market change given a significantly reduced fee for servicing, and is there enough incentive for servicers to stay in the business?
As we ask ourselves the question above, the impacts appear to be significant and may vary based upon the following considerations.
- Large versus Small
- Economies of Scale or Barriers of Entry
- Reduction in Capital Requirements, Reduced MSR balance sheet exposure
- Decrease in use of Financial Hedging Products
- Reconfiguration of Modeling Requirements
- Impact to the TBA market
- Potential for Adverse Selection
- Reduction in Earnings volatility
- Redistribution of Market Loss exposure
- Reduction of GSE loss risk exposure as guarantor insurance helps to offset future losses.
Incentive and disincentive policies to redistribute the industry; awarding those outperforming while penalizing those with below average performance. Ideally, this should lead to overall performance enhancement.
Consolidation of Residential Servicers
At approximately 64%, the top five residential servicers have steadily grown their market share in recent years. In MIAC’s opinion, one substantial reason for the increased market share has to do with volume based guarantor fee structures. Small to middle tier servicers are not able to compete as effectively on price due to the preferential guarantee fees assigned to higher volume producers. The combination of higher guarantee fees and more prohibitive economies of scale oftentimes make it economically preferential for smaller servicers to sell to one or all of the aggregators opposed to retaining it themselves. Following, we will go into greater detail regarding a potential solution.
A Credit Risk Tiered Guarantee Fee
A risk tiered guarantor credit structure would appear to be the most fair and therefore the best approach to changing the current service fee structure. (A more detailed discussion of this concept is described later in this article.) Based on the market acceptance of this risked tiered credit structure, consumers would receive market note rates determined by their credit and risk profiles. The ranking of these credit and risk profiles will require standard guidelines to ensure that consumers receive mortgage market rates that accurately reflect borrower risk. Higher risk and non-qualified mortgages will incur higher guarantor fees with the borrower sharing at least some of this cost in the form of above market mortgage rates. The higher guarantor fees would be compensation for the increased credit risk as would the higher note rate premium required for these higher risk loans.
Consumer Impacts
In a proper risk tiered guarantor credit structure, the higher costs will properly reflect the credit risk of the mortgagor. Higher note rates, reflective of credit will dictate higher reserves. Profit margins previously earned by originators charging higher rates and premium fees will now be passed through to the guarantor entity (i.e., GSE’s or Private Securitization).
Some consumers also will benefit from the Dodd-Frank Act, which promotes the establishment of specific exemptions from risk retention requirements for “qualified residential mortgages” (QRMs) using criteria that demonstrates lower collateral risk characteristics that should result in better delinquency performance and lower default risk. Such (QRM) characteristics as fully documented underwriting, higher FICO scores, lower loan to values (LTV’s) and the use of credit enhancements should result in minimal market rate impact for qualified (QRM) borrowers.
Investor Impacts
The overall investor impact should be positive. The risk based guarantor insurance premium (RBGIP) will act as a loss reserve escrow account from which to draw from for future losses and for the additional servicing requirements for the NPL loans.
As our tables below suggest, the risk based guarantor insurance premium (RBGIP) should be reflective of the projected loss exposure over the lifetime of the loan. As empirical research suggests, various loan level attributes can be utilized to develop a more robust risk tiering that can provide better approximation for the perceived loss exposure. For instance, a 750 Fico, 75 LTV, at market rate conforming loan may incur a lower cumulative default and be considered low risk, thereby incorporating the smallest risk based guarantor insurance premium, 12.5 basis points in our example.
On the opposite side of the credit and risk spectrum, a high risk loan such as subprime type collateral that incorporates a low Fico, high leverage, and high projected defaults should utilize a high risk based guarantor insurance premium (RBGIP). In this example, we assume a 40% cumulative default, a 20% cumulative loss and show various (RBGIP) that may properly reflect the added risk and loss exposure. There are multiple iterations that can be incorporated to arrive at an acceptable risk based reserve. Changes to the default and loss severity projections and to the timing of the losses need to be considered. In some instances, the performing low risk loans paying a 12.5 basis point (12.5 bps) risk based guarantor insurance fee premium (RBGIP) may subsidize for the high risk loans that may have larger risk exposure than the risk based guarantor insurance premiums that are being charged to them. Exploring the proper allocation for the (RBGIP) can be discussed at a later time.
As the tables suggest, the accumulated risk based guarantor fees will aid the investor, classified as the GSE’s or the Private Label Originator as they will accumulate more credit reserves to offset future collateral losses. First line losses should be covered by this increased insurance fund reserve. This reserve (insurance) fund will benefit from the originator having more skin in the game, thereby reducing investor losses. As borrower credit profiles decline, more reserves in the form of increased guarantor fees may be set aside to allow for the increased projected losses.
Table 1 provides the cumulative loss reserves for a $200,000 loan balance using a 15% CPR and various cumulative default projections. These figures are the projected basis point values for the various scenarios.
For instance, based on a 3% cumulative default and a 12.5 basis point risk based guarantor insurance fee premium (RBGIP), the investor (GSE’s or Private Label Investor) will be able to withstand 67 basis points of losses. The Stressed scenario that incorporates a 40% cumulative default and 300 basis points of (RBGIP) can withstand 1,235 basis points of losses. On the $200,000 loan, the accumulated (RBGIP) would range from $1,342 to $24,697 of cumulative loss coverage based on these scenarios.
In time, this may result in increased investor demand, resulting in tighter spreads and lower yield requirements due to the enhanced risk profile of this new structure. Some may argue that a reduced service fee may have an impact on prepayment speeds. Adverse selections or higher amounts of prepayment churning are often topics of discussion from investors. MIAC does not see material prepayment risk resulting from a reduced service fee.
Servicer Impacts
Smaller servicers that did not have performance issues, namely delinquency and advance issues, will have a significant reduction in revenue income derived from the reduced service fee. Given their higher servicing cost structure, higher market yield requirements, and higher capital costs, the smaller servicer may find that reduced MSRs no longer cover the risk in owning mortgage servicing rights. Prepayment volatility will still be a factor, albeit it will have a smaller impact to overall values given the reduced service fee.
Below, MIAC demonstrates the impact to the servicer with various cost structures, default projections, and prepayment scenarios. In our servicing cost example, refer to the $100 marginal cost projection as reference for a small servicer.
Additionally, the smaller servicer often does not service the loan internally on their own servicing platform and may use a sub-servicer. This frequently results in revenue streams exclusive of service fee income (namely, late fee, ancillary income, and float income) being partially or fully passed through to the subservicer. Higher expenses (servicing costs, funding costs) with less revenue offset (late fee, float, and ancillary income) may result in fewer lenders being able to continue holding MSRs. The smaller servicer may hold a smaller portion of MSRs and/or may sell the MSRs servicing released (AOT).
The end result is likely to be fewer servicers. The risk is that all servicing is eventually congregated with a few, large investors that have robust enough cross selling operations that make servicing for reduced revenue possible to stay in the servicing business. A business line that has for years been driven by economies of scale will be become even more so. Additionally, most lenders consider the home mortgage a central part of the customer relationship from which many other mutually profitable relationships are formed (e.g., checking, savings, money market, credit card accounts, etc…) Most lenders agree that not managing the mortgage servicing relationship frequently results in losing the customer altogether to larger servicing institutions over time. Small originators will have diminished customer relationships. Additionally, servicing congregated at only a handful of servicers puts taxpayers at further risk of investors becoming “Too Big to Fail”. Whether the eight-hundred pound gorilla investor is public or private, their value and risk to the overall economy will be increased and undeniable.
Large servicers should experience more positives than negatives. On the assumption that large servicers (namely banks) have lower cost for capital, lower servicing costs due to economies of scale, and in general lower yield requirements, these large servicers should still be able to make a profit on the performing MSRs, (albeit at a reduced service fee revenue projection). However, late fee income, float income, and ancillary income resulting in fewer cross selling opportunities will be disadvantage the smaller servicer.
In addition, the large servicers tend to have hedging programs in place to manage the interest rate prepayment exposure and earnings volatility. By having a smaller service fee and hence MSR financial exposure, the large servicer may not require as active a hedging strategy and profile, thereby resulting in decreased hedging costs and overall expenses. Some of the reduction in revenue from the reduced servicing fee may be offset by these lower expenses (hedging costs).
Additionally, with a lower value assigned to the MSRs, the mortgage servicer’s balance sheet related to MSRs would be reduced, a positive for Tier1 regulatory capital requirements under Basel III.
In our servicing cost example, refer to the $50 marginal cost projection as reference.
Overall impact to the servicing market of a reduced service fee and the use of a risk-based tiered guarantor fee may promote higher quality originations. Higher risk collateral that would already incorporate higher mortgage market rates would incur a higher guarantor fee structure resulting in higher note rates (premium coupon) product. From a servicing standpoint, this high premium coupon product would have lower servicing values for at market and below market rate mortgage servicing rights.
Modeling the correct guarantor fee to correspond to the credit loss projection will become very important so as not to overcharge or undercharge the borrower for the perceived risk exposure to delinquencies and losses. MSR values will be attributed to the projected performance and collateral attributes of the servicing asset.
Originator Impacts
The quality of product being originated will drive resulting impacts to mortgage originators. One suggestion would be to apply lower guarantor fees and minimal changes to mortgage rates for (QRM) approved borrowers. The originator business model would remain largely the same; the GSE (or Private Label) Originators’ would be able to more effectively manage risk and potential losses through this risk based guarantor fee approach.
Production Impacts
There would be a redistribution of originated product.
- The Tiered structure might work simply as follows: Higher credit risks entail higher note rates.
- This would naturally create a natural disincentive to originating high risk product, as guarantor fee insurance premiums will increase based on risk profile.
- Regarding Current Production, the tier structure might be as follows:
- Premium Coupon = High Risk = Higher Guarantor premiums set aside.
- Base Coupon= Base Risk = Base Guarantor premiums applied.
- Discount Coupon = Low Risk = Pristine Collateral = Lower Guarantor premiums applied = may be lower than amount currently set aside.
The tiered structure could follow the MPF program concept where the originator has a financial stake in the loan (“Skin in the Game”). Based on the credit /risk profile, variable guarantor fees will be paid to this insurance fund reserve account. Like the MPF concept, should the originator/servicer perform better than the benchmark set during a stated period of time, some of the added cushion can be released back to the servicer as an income incentive. This concept can be modeled based on timing of release of funds Net Present Valued. (NPV).
For example, if the target is for a 5% loss reserve by year 4, and by year 4 the client has experienced 1% of losses yet they have set aside 5% of loss reserves, the client will have the option of pulling out up to 50% of the allotted reserves in year 4. Three (3) % of the total loss reserves will be accounted for upon this release. (5%-1%)*50% + 1% actual losses. The company will incur this remaining 2% income payout as continued incentive for their performance.
For the opposite side of the spectrum, if the client is underperforming the target benchmarks, they will be required to increase their reserve requirements on their current production or pay up the shortfalls.
Financial Impacts
Overall dollar denominated value from the servicing fee reduction will result in decreased MSR dollar denominated values. Although servicing fee multiples may not dramatically decline as our valuation examples state, the revenue streams from these service fee cash flows will be dramatically reduced. For example, the service fee income on a $200,000 principal balance loan @ 25 basis points (bps) is $500.00 per year. Should the servicing fee utilize 12.5 bps, the service fee revenue would be cut in half to $250.00 per year ($200,000 * 0.125%). Consider the chart below.
Firms may want to explore opportunities to amend MSR pricing and reduce costs while assessing the profitability of their existing business models. In some instances, companies may consider exit strategies or curtail the amount of business related to the MSR asset. This is particularly worrisome for marginally profitable lending and servicing businesses and capital market and trading groups.
In addition, financial reporting may require the separation of Legacy Loans and New Production under the new service fee structure. Companies would need a separate valuation analysis for both legacy and the new structure. The older (legacy loans) would be in runoff mode estimated at over $5 trillion UPB.
Operational Impacts
Some operational systems may need to be reconfigured to allow for the new service fee structure. In general, isolating the performing and non performing non legacy loans would be required. Additional operational oversight would probably be necessary on the non performing product. Questions concerning the impact of a loan moving in and out of various stages of delinquencies and whether the loan would reside with the original servicer or move to a separate non performing group (internal or external) will need to be addressed. In addition, the operational requirements to transfer the delinquent servicing may require some changes.
Accounting Impacts
A reduction of the mortgage servicing fee would result in a decrease in the size of mortgage servicing rights (MSRs) now recognized on a mortgage servicer's balance sheet. The service fee reduction may reduce the potentially sharp swings in the value of MSRs due to interest-rate risk exposures and the servicers' earnings volatility. Due to the capital intensive nature of MSRs, a reduction of the service fee may result in the reduction of servicer’s effective leverage.
Under current accounting rules, mortgage originators recognize a separate servicing asset upon the securitization or sale of mortgages, servicing-retained. The two assets that are created include a debt instrument and the mortgage servicing rights (MSR). MSRs are a series of cash flows retained by the servicer representing future fees for collecting and disbursing the monthly interest payments to the mortgage investors, the collection and payment of property taxes, insurance and the handling of delinquent loans. The MSR servicing asset represents the present value of net future cash flows that will accrue to the mortgage servicer over the life of the loan.
Financial Reporting
New MSRs will be booked at a lower dollar denominated value given that the servicing fee has been reduced. Tier 1 capital requirements for the non legacy MSRs will probably be reduced given reduced capital exposure to the MSRs. It is possible that the legacy loans may need to be grandfathered; however, non legacy MSRs will be in runoff mode and their exposure will decline with time (prepayments).
Financial Systems’ Software
Changes to modeling assumptions would be assumed. MIAC has the software capability at the present time to handle the structural changes to the servicing fee.
Performance Impacts
With the implementation of a successful risk based guarantor credit tier structure, credit performance should improve as high risk product will require higher guarantor fees and result in a disincentive in originating high risk product in excess. Given that pristine collateral will require minimal guarantor fee (insurance) coverage, the market may evolve toward more effective market-based pricing.
Model Values
MSR Values for legacy MSR portfolios may or may not be impacted. An interesting question is the impact on liquidity should Basel III be implemented. The capital requirements on the legacy MSR portfolio may warrant MSR sales that can be impacted based on market conditions.
Values for new production MSRs using the revised servicing fee structures would result in lower dollar denominated servicing values given the lower revenue streams incurred from the service fee reduction.
MIAC Takeaways
We have provided our viewpoints concerning the impact of a change in servicing fees on the current market. It seems clear that a risk based guarantor insurance structure can be effective in providing the proper capital reserves to handle losses due to delinquency performance. Inadequate capital in the system left financial institutions unprepared to absorb losses and damaged the financial system and the confidence in the markets. The mortgage servicing market was not prepared to handle the perfect storm of a capital market shutdown, reduced liquidity, portfolio performance deterioration, and higher expenses against a backdrop of declining servicing revenues and home price deterioration.
Further discussions will ensue on how viable service fee and guarantor fee structures (that can help level the playing field for participants and balance the risk and reward profile) will be priced in the markets in an efficient and effective manner. MIAC looks forward to working closely with our clients in providing further analysis and solutions for any changes in the current service fee structure.
Download .pdf
[BACK TO MAIN PAGE]