Mission & Metrics Finance Training for US Credit Program Professionals

Transcription

1 MIT Golub Center for Finance and Policy presents: Mission & Metrics Finance Training for US Credit Program Professionals December 12-13, 2016 This course is made possible by a generous gift of Mary and Warren Naphtal to the GCFP Naphtal Education Initiative Fund 1

2 U.S. government is world s largest financial institution Graphic from Politico, The (Real) Bank of America, by Michael Grunwald, Jan/Feb

3 Diverse set of credit & insurance activities Credit & Insurance Outstanding, 2014 (trillions of $) $7 $6 $5 $4 $3 $2 $1 $0 Deposit Insurance (FDIC) Fannie and Freddie Pension Guarantees (PBGC) Tradi onal Credit Programs Home Loan Banks (FHLBs) Farm Credit (FCS) 3 Excludes social insurance programs and Federal Reserve activities

4 Dollars (Billions) Loans across government show continuing growth Total Federal Loans Outstanding (Direct and Guaranteed) by Category: Housing Education Small Business and Farming International Lending Energy 4

5 Why mission & metrics? Mission Enabling agency mission accomplishment through credit programs Key considerations: Metrics 5 Be ever mindful of using taxpayer dollars (and putting them at risk) Determine, on ongoing basis, whether program is working as intended Seek ways to measure & improve program performance Collect & analyze data to gain understanding of program costs & benefits Key considerations: Embrace analytics to drive innovation, measure & manage performance & motivate staff Take occasional step back to gain big picture insights Remain passionate about agency mission, but dispassionate about programs being used

6 Course objectives and structure Gain deeper understanding of the following: Program objectives - Understanding mission and how best to achieve it Considerations in program and product design Finance principles involved in assessing and forecasting credit costs Discounted cash-flow modeling for credit programs Economic cost drivers and relation to budgetary subsidy measures Managing operating costs and credit risk Students will consider: Economic rationale for government credit support Considerations in the use of credit to achieve mission Methods for measuring the costs of credit activities 6 Management issues flowing from program design & cost

7 Course objectives and structure Format: 7 Lectures and participant discussions Your voice is critical! Instructors: Deborah Lucas (dlucas@mit.edu) Doug Criscitello (dcris@mit.edu) Charles Tansey (cdtansey@mit.edu) Roman Iwachiw (iwachiw@ficonsulting.com) Complementary perspectives: Insights from financial economics as taught in b-schools and used by private sector financial practitioners Managerial and budgetary considerations and best practices Practical tools from the front lines of lending in the private, non-profit and public sectors

8 MIT Golub Center for Finance and Policy Mission is to produce and disseminate: Original, cross-disciplinary and non-partisan research at intersection of finance and policy that will lead to improved decision-making by policymakers and regulators Innovative educational materials and curricula that will make state-ofthe-art financial tools relevant and accessible to students of public policy, employees of public institutions and policymakers Please visit us at GCFP.MIT.edu for more information on the GCFP 8

9 Session 1 Dec. 12, 1:00-3:00pm Program Objectives - Understanding Mission and How Best to Achieve It 9

10 Economic rationale for government credit support The story (famous paper by Joseph Stiglitz and Andrew Weiss Rationing in Markets with Imperfect Information, AER 1981) No special role for gov t credit in a perfect market But credit markets are not perfect. Asymmetric information between borrowers and lenders and other contracting problems lead to: Adverse selection (higher rates attract the riskiest borrowers) Moral hazard (debt induces risk-taking behavior) Those problems limit banks willingness to lend at any price It is beneficial for the gov t to fill the resulting credit gap when access to credit leads to socially desirable outcomes whose value exceeds the cost to taxpayers of the credit support 10

11 Economic rationale for government credit support Interest rate supply demand 11 Quantity of loans In a perfect market rates clear markets Gov t can help borrowers with grant to help pay interest

12 Economic rationale for government credit support Interest rate supply (actual) supply in perfect market demand 12 credit gap Quantity of loans Higher rates => lower borrower quality No market-clearing interest rate; credit rationing

13 Economic rationale for government credit support Takeaways Theory points to when credit support will be most useful: Severe lack of information about borrower quality Greatest social benefits from providing access to credit Classic example is students loans no credit or employment history private and social gains to having a more educated populace Theory also highlights that greater availability of funds can be as or more important for recipients than lower rates and fees 13

14 Economic rationale for government credit support Takeaways The gov t can increase credit availability because, unlike banks, it does not have a breakeven constraint That is, gov t can offer credit on subsidized terms This can benefit society when expanding access to credit is the best way to achieve a policy goal This logic also implies that there is generally an economic subsidy when the gov t extends credit. Otherwise private lenders would step in to compete to capture the profit 14

15 Drawbacks of credit support Shortcomings as a policy tool: Opaque (both costs & benefits) compared to most other types of spending Can distort capital allocation and crowd out private capital Disguised industrial policy? Can encourage excessive indebtedness Incentives for excessive risk-taking with systemic consequences e.g., credit policies may have contributed to the 2007 financial crisis Often target-inefficient Carefully thought out program objectives and designs, and metrics for performance evaluation, can mitigate these problems 15

16 Example: How a credit guarantee can increase borrowing volume at a modest subsidy cost 16 Percent guaranteed Source: Credit Policy As Fiscal Policy Deborah Lucas, Brookings Papers on Economic Activity, 2016

17 Example: Estimated fiscal stimulus from federal credit programs in aftermath of financial crisis In 2010, U.S. citizens borrowed $1.6 trillion through federal credit programs The spending resulting from that incremental borrowing provided roughly $344 billion of stimulus in 2010 by filling the credit gap Estimate is from Credit Policy As Fiscal Policy Deborah Lucas, Brookings Papers on Economic Activity, 2016 Similar stimulus size as the American Recovery and Reinvestment Act Cheap stimulus: Estimated fair value cost of $71 billion; reported budgetary cost of -$12 billion 17

18 Mission Mission statement = purpose of organization & its reason for existing Guides actions Spells out central goals Provides a path to guide decision-making Provides the framework or context within which strategies are formulated Among US departments & agencies, only Treasury makes any reference to ensuring appropriate use of taxpayer dollars Manage the U.S. Government s finances and resources effectively Consider working every day as if the following clause appears at front of your agency s mission: As constant stewards of taxpayer dollars, we seek to [insert mission] Given large exposures across loan programs, mindset is absolutely critical for federal credit professionals 18

19 Loan program justification varies SBA Education HUD USDA HUD Energy Treasury - CDFI Fund Agency Program Small Business Investment Companies (Venture Capital) Federal Student Aid FHA Rural Development FHA Clean Energy CDFI Fund Primary reason Financial Policy Financial Financial Financial Policy Policy Who championed the problem? SBIC Trade Group Public Congress Public Congress Administration Administration The debenture structure was inappropriate for funding early stage VC deals Accelerating college costs, decreasing state support, abuses in forprofit entities, flawed private sector lending prompted takeover Seniors with equity in their homes but inadequate cash can use the equity to boost cash available for living and other expenses Limited banking options for low and very low income residents in rural areas make hard to own a home, contributing to disinvestment Following the Debt Crisis there was a need to help defaulted borrowers to get current, renegotiate their mortgage debt, and avoid foreclosure There was a policy need to spur development of clean energy products through commercialization of innovative ideas CDFIs with excellent financial performance cannot get long term funding due to market unfamiliarity. Fed bridges the knowledge gap but wants no risk Product Participating Securities Direct Student Loans Reverse Mortgage Insurance 502 Direct ELP Title 17 CDFI Bond Guarantee Program 19

20 $1B, 1-year EHLP Key challenges Managing demand -- Jan M unemployed & 4.5M borrowers delinquent but $1B EHLP can fund loans to about 40,000 borrowers Cost -- Using judgment gained from data from a similar program operated at the state level, estimated only 3% repayment rate plus admin costs of maybe percent, expected costs are about $1.10/dollar loaned Implementation -- Relied heavily on contractors to deliver program, with no time to test approach before go live Workload Sought to minimize administrative expenses but needed to be nimble enough to apply resources as needed Fraud prevention No traditional loan underwriting; once in, borrowers required to certify annually whether remain eligible for assistance (if hired or otherwise become ineligible for EHLP, required to notify HUD) Headline risk Given above, needed plan to coordinate inquiries from the Hill, press, OMB, OIG, and other interested parties Metrics How to measure success? 20

21 Recent focus on data-driven decisionmaking Emphasis on value of data, agency performance and results to achieve a smarter and more efficient government Example such as GPRAMA, DATA Act, Accountable Govt Initiative aim at improving agency performance & using data to make funding decisions Cost information is critical -- informs management decisions & contributes to improved planning, implementation, analysis of programs Benefits (or lack thereof) must also be measured Particularly relevant for credit agencies 21

22 and on risk management tools for credit agencies Enhanced lender compliance (counter-party risk) monitoring & increased supervision/enforcement actions Use of analytics to measure risk & inform decision-making Enhanced risk assessment frameworks composite views of risk (e.g., loan performance, financial condition, regulatory compliance) Growing number of risk management organizations leading practices & capabilities from former financial services professionals now at credit agencies Better use of enterprise-wide risk management (ERM) & increasingly risk-aware cultures Increasing # of credit agencies with CROs & independent risk organizations and governance structures (e.g., risk committees) 22

23 A-129 guidance Program Reviews & Evidence Building Frequent & specific Define, quantify & justify Program Management: New Content Covers organizational structure, risk management & metrics Data-Driven Decision Making Federal Credit Policy Council 23

24 Goal of continual program improvement Need for constant program monitoring & refinement Measure impact thru periodic strategic reviews Are public policy objectives being met? Does need for program persist? Are there ways to achieve goals at less cost/risk to taxpayers? 24

25 A-129 checklist What are the objectives of this program? Why is federal credit assistance the best means to achieve these objectives? What are the existing and potential private sources of credit? Why do private sources of financing need to be supplemented and subsidized? What is the scope of the program, the amount of federal credit assistance estimated necessary to efficiently meet the intended objectives, and reasons why a federal credit subsidy is the most efficient way of providing assistance? What are the benefits expected from the program? What are the elements of design which encourage and supplement private lending activity, such that private lending is displaced to the smallest degree possible by agency programs? How does the program s structure encourages risk sharing with private lenders? What are fees and interest rates, and how they are consistent with credit policies? Do terms and conditions protect the Government s interest? What are the methods used to evaluate the program and the results of evaluations that have been made, including supporting analyses or other evidence? Do proposed changes improve the efficiency and effectiveness of the program? What is the plan to measure impact? But mostly, what s the problem we re trying to solve? 25

26 Defining the Market and the Mission What are the Primary rules for providing federal credit? Define the perceived credit gap and fill only the credit gap as defined Make sure the borrowers or the projects are viable, or if the risk is exceptional, that it is defined as such Do not compete with (responsible) private sector providers Pass the credit gap back to the private sector as soon as private sector lenders show interest in filling it Key Steps 1. Objectives: What, When, Who, Why, How. 2. Performance: Managing cost, risk and quality 3. Outcomes: Quantified and tied to the Objectives 26

27 Justification for Intervention: Key Product Indicators OUTCOMES HOME MORTGAGES Conventional Credit Gap SMALL BUSINESSES Conventional Credit Gap Potential Annual Volume $300 Billion $25 Billion #500,000 #20,000 Deal Size $350,000 $250,000 $1,000,000 $150,000 Credit History Good, > 5 yrs old OK Pay History, credit score >160, >2 yrs old Term 30 years 30 years fixed 5 years, term loan 7 years working capital Capacity to Pay D/I = 33% D/I =40% DSC 1.5x DSC 1.2x Collateral LTV = 95% LTV = 99% 1.25x 1.0x Location Stable Market, urban, suburban Declining/Distressed or rural Stable Market, urban, suburban Declining/Distressed or rural Regulators No problem Limit exposure No problem Limit exposure, Raise capital allocation 27

28 Objectives, Performance and Outcomes (1) THE KEY STRATEGIC QUESTIONS EXAMPLES OF THE ANSWERS What are we providing? Who is the beneficiary? What is the size of the problem? Who are the partners? What are the Agency resources available? 100% guarantee of a 1st lien mortgage for first time homebuyer for up to 30 years at a fixed rate for loans not exceeding $250,000. Debt to Income not to exceed 40%, downpayment not to be less than 1%, fees can be financed. Homebuyer counseling is required. Fixed rate at the prime mortgage rate. Up to 80% AMI, income averaging $32,000, credit score not less than 620. Priority for census tracts and tracts where rental housing is not available or exceeds 50% of household income. Census and market data indicate 1million households in this affordability gap. Depositories do not serve this public due to regulatory constraints. The pricing from nondepositories is excessive. Community banks, credit unions, approved non-profits and consumer lenders. Target: 400 participating lenders. 12 FTE including 3 underwriters, 3 monitoring specialists and 2 workout specialists. Staff projected to double over the 10 year period to be sourced from an existing programs that is being phased out. 28

29 Objectives, Performance and Outcomes (2) KEY COMPONENTS OF PERFORMANCE KEY PERFORMANCE INDICATORS How do we manage credit risk? The chief variable is credit risk. Credit scoring to be used for all borrowers and predictive ortfolio scoring for participating lenders. Quarterly borrower credit score updates and lender assessments. Production benchmarks comparable to private sector. System of clear graduated remedies that strictly enforces lender standards, and eliminates outliers. How do we manage operating costs? How do we manage the borrowers? How do we manage the lenders? How do we monitor changes in the market and how do we adjust for them? Automated application and data capture, storage and management enable us to generate volume prudently at private sector cost. Monthly reporting on historical/ budget/ actuals on comprehensive operating data as enables us to detect deviations from plan and production benchmarks. Regular servicing performed by the lending partners. We contact the client directly after 60 days delinquent. There is an annual survey of all borrowers. We perform predictive analysis of all lenders annually, and perform credit audits and site visits on the outliers based on the system of graduated remedies. Annual customer and lending partner surveys with interpretive analysis. Volume, credit risk and pricing analysis in comparison to market trends. In the event of a downturn and the prospect of higher losses, we pass the increased cost of risk through to the borrower. If it exceeds the 40% Debt to Income target we reduce lending volume. In the event that usage drops below 500 originations per year we terminate the program. 29

30 Objectives Performance and Outcomes (3) OUTCOMES Credit Risk Program Costs Lenders Borrowers The average credit score for this year's new loans was 640, 10 points lower than last year's. Portfolio delinquency is at 4.9%, higher than plan and the prior year. Credit losses of 1.8% are also higher than prior year and plan. The number of defaults within 18 months of origination has declined 18% year to year. We are exceeding operating benchmarks: (1) backlogs have dropped from 8 days to 6.5 days; (2) loans per underwriting FTE exceed 1,000; and (3) modification response time has dropped from 9 days to 8.2 days. 420 lenders nationally are signed up for the program. We are adding an additional 30 this year net of 10 year that were removed from the program. Average volume per lender per year is 11 transactions. Average portfolio risk has risen: this year's loan volume carried lower credit scores and weakness in the economy has caused a rise in delinquency and credit loss. We were able to help 9,000 more households lower the cost of housing under 40% of income. We helped reduce the demand and price for affordable rental housing in targeted census tracts by 3%. We helped reduce the eviction percentage in the target population by 10% (on a pro rata basis). On average, credit scores have risen within 3 years of origination under the program. 30