Highest CPI since 2008, Strong FinTech Earnings, PeerIQ’s Valuation Methodology

by | Aug 12, 2018 | Blog | 0 comments

Greetings,

US Consumer Price Inflation (CPI) rose by 2.9% YoY in July. Core CPI, which excludes volatile food and energy prices, rose by 2.4% YoY, the highest reading since September 2008. Rising inflation is squeezing real wage growth, which has averaged just over 2% this year. You can read our previous deep-dive into various inflation measures and their impacts on Fed policy here.

In this week’s newsletter we analyze FinTech’s earnings and introduce PeerIQ’s valuation methodology. We identify the top 5 questions to ask to know whether your whole-loan valuation makes sense and is regulatorily compliant.

Also, thank you for joining PeerIQ’s Webinar on OCC’s New FinTech Charter – What Does it Mean for Lenders and Investors? The webinar’s reception was a terrific success, and we want to thank our panelists again for their contributions.

The webinar delved into the Treasury’s report on FinTech, the OCC’s guidance on the charter, risk management and supervisory standards, and whether FinTechs should consider an application.

Download the Presentation
Watch the Replay

PeerIQ’s analytics tools can help to comply with the Capital & Liquidity and Risk Management requirements of the OCC’s charter. FinTechs that can supercharge their risk management, reporting, and analytics to comply with the charter’s guidance have a competitive advantage. Reach out to learn more.

 

Strong Fintech Earnings

LendingClub, OneMain, OnDeck, GreenSky and SoFi reported earnings recently. We summarize them below:

LendingClub delivered record quarterly revenues of $177 Mn, an increase of 27% YoY. Net loss also increased significantly from $25 Mn to $61 Mn YoY.

  • Originations grew by 31% YoY to $2.8 Bn, the highest quarterly originations at LC.
  • Net interest income was offset by fair value adjustments on loans of -$27 Mn.
  • Structured program generated revenues of $5.6 Mn.
  • Projected returns across grades increased by between 8 bps and 40 bps QoQ across grades.

Revenue at OneMain grew by 17% YoY to $905 Mn while net income fell by 83% YoY to $7 Mn.

  • Receivables grew by 11% YoY to $15.4 Bn.
  • Provision for loan losses increased by 12% YoY to $261 Mn, in line with loan growth.
  • Net charge-off rate decreased slightly YoY from 6.9% to 6.6% YoY.

OnDeck’s revenue grew by 19% YoY to $50 Mn and net income grew to $5.8 Mn from a loss of $1.5 Mn YoY.

  • Loans grew by 8% YoY to $1.1 Bn, while originations grew by 27% YoY to $587 Mn.
  • Provision for loan losses increased by 2% YoY to $33 Mn.
  • Net charge-off rate decreased significantly from 18.6% to 11.2% YoY.

GreenSky reported its first earnings as a public company this quarter. Revenues grew by 28% YoY to $106 Mn, while net income was $41 Mn.

SoFi reportedly lost $200 Mn in the 2nd quarter driven by lower valuations on legacy loans due to rising rates.

We will track all bank, credit card issuer and fintech earnings in our Lending Earnings Insights Tracker. The previous tracker can be found here.

 

PeerIQ Valuations – Approach and Overview

Over the next few weeks we will take you through PeerIQ’s approach and methodology for valuing whole-loans. This week we focus on the top 5 questions to ask to know if your valuation makes sense.

As private credit markets grow, there is increased regulatory scrutiny on whole loan valuations. As mentioned in Asset-Backed Alert recently, the SEC will soon require investment advisors to have a more robust whole-loan valuation framework. Failure to comply with such regulations could invite enforcement actions which hit a record last year, and where actions were concentrated in private credit and pricing of illiquid collateral in particular.

Valuation of a portfolio of unsecured personal loans is inherently difficult as there are generally no market-observable prices to indicate where comparable portfolios are trading. Additionally, Interest rates curves and credit spreads are constantly changing which makes estimating a discount factor challenging. Also, portfolios must be valued at an individual loan level which requires having loan-level cash-flow projections using individual CPR and CDR curves based on clean (sometimes daily) data.

Estimating loan-level curves in the absence of a large observable dataset is impossible, making whole-loan valuation a big-data problem that can be solved only by using the right technology.

Regulatory Framework for Valuation

The Financial Accounting Standards Board (FASB) provided guidance for fair value measurements and disclosures in ASC Topic 820, including the Fair Value Hierarchy across three levels:

  • Level 1 assets have actively traded markets; and valuation inputs are direct quoted prices for identical assets or liabilities. (Think stock-prices for large-cap stocks for example.)
  • Level 2 assets include valuation inputs other than quoted prices that are observable for the asset, either directly or indirectly. Examples of Level 2 assets are privately placed bonds whose value is derived from a similar bond that is publicly traded, corporate bonds, collateralized mortgage and debt obligations, and high-yield debt securities.
  • Level 3 assets are at the most unobservable level and require many assumptions and estimates. Since a two-way market is not available, fair value is often based on models in which there are few, if any, external observations. Examples include nonpublic private equity investments, retained interest in securitizations, and mortgage servicing rights.

Under ASC 820 guidelines, the fair value of an asset reflects the exit price that would occur in an orderly market.

In addition to the ASC 820 guidelines, lenders and asset managers must comply with the Current Expected Credit Losses (CECL) standards latest by December 2020. CECL requires estimates of potential losses over the life of all originated or purchased loans. Estimating losses over the lifetime of the loans requires significant enhancements to current valuation models to get accurate cashflow projections.

Does your Valuation Make Sense?

Here are the top 5 questions to ask to understand whether your loan valuation makes sense:

1. Are there conflicts of interest?

A primary requirement for a valuation agent is independence. When an asset manager performs valuations on its own portfolios, there are conflicts of interest as the manager earns performance fees based on their own self-assessment.

We have seen regulatory action against funds like Platinum Partners relating to overvalued loans due to this conflict of interest. Such problems can be avoided with the help of an independent third-party valuation agent.

A “shadow valuation” agent can also be extremely useful in ensuring that the prices provided by the primary valuation agent are accurate. A shadow valuation involves getting a second (or even a third) set of prices for loans in the portfolio and checking for significant valuation discrepancies among them.

2. Are seasoned loans valued at a discount to unpaid principal balance?

Valuing seasoned loans accurately is one of the biggest challenges for asset managers. Market observations on such portfolios are exceedingly rare, so asset managers sometimes incorrectly holding current/performing seasoned loans on their books at par, with market value equal to the unpaid principal balance.

This approach is incorrect as the schedule of cashflows for a loan changes as the loan seasons. For example, for unsecured personal loans, charge-offs can meet or exceed interest received for several months – in sharp contrast to the performance of a newly minted loan.  Unsecured personal loans display delinquency profiles that peak at roughly a third of the way into the life of the loan. Adjusting the loan’s price for the probability of default at any point leads to prices below par as newly issued loans season, and a pull-to-par effect as the loan approaches maturity.

The chart below shows the illustrative price profile on a sample unsecured personal loan.

Source: PeerIQ

3. Are the impacts of changes in interest rates and credit spreads considered?

Rising interest rates and widening credit spreads increase the yields, and consequently, the discount factors of personal loans. Failure to do so will lead to a portfolio value that is inconsistent with changes in the market.

4. Are you applying a fair value approach, or applying any of the following faulty methodologies?

Below are some faulty methodologies that are currently being used:

Amortized Cost: This approach values a loan at its outstanding balance at purchase price plus accrued interest. It may overstate loan price during the early period of the loan’s life, as it does not account for loan status (e.g., delinquency). (Geek Note: Amortized cost method makes sense only for banks where loans are held-to-maturity, AND there is no money entering or leaving the fund. Even in this case, the goal of amortization is concerned with accounting /cost allocation rather than fair value.)

Haircut Matrix: Loans are valued at outstanding balance plus accrued interest, with haircuts applied to loans based on their stage in the delinquency queue. The size of the haircut is calibrated to historical loan performance.

The haircut-matrix approach improves on an amortized cost approach by incorporating loan status, yet still suffers from major deficiencies ignoring A) changes in a borrower’s credit profile at loan level, B) seasoning of loans, and C) credit spread and interest rate risk premium, and D) a forward-looking view on cashflows.

Loan Loss Provisions: Banks holding loans in the hold-to-maturity book create a provision or accrued liability based on expected losses of the pool. This approach is not applicable to hedge funds offering liquidity as it leads to “stale” and smooth loan prices that do not reflect fair value.

5. Is poor performance being masked with new originations?

As personal loans usually reach their peak delinquency roughly one-third of the way into their lives, performance on new originations is significantly better than that on seasoned pools. If the volume of new originations outweighs seasoned loans, then the returns of the portfolio may be artificially inflated. Once new origination stops, and the pool starts seasoning, returns will drop, and loan performance will deteriorate. Properly controlling for vintage effects is critical to maintaining the integrity of valuations.

Key Principles of Robust Valuation Approach

The PeerIQ pricing and valuation framework adheres to the following key principles:

Regulatory Compliance: The valuation approach and output should comply with ASC 820 and CECL guidelines with a robust and defensible methodology.

Transparency: A repeatable, documented methodology, with exposition of inputs and other assumptions.

Fair value: If the valuation inputs are not readily observable, ASC 820 requires a derived fair value based on a consistent modeling framework that factors in both unobservable and observable valuation inputs from similar assets.

Sensitive to major risk factors: The pricing framework accounts for risk factors including default risk, prepayment risk, and credit spread risk.

Forward looking: Loan valuations are driven off expected future cashflows, with such cashflows a function of the borrower’s attributes, macro conditions, loan performance, and seasoning.

Loan-level: The pricing framework operates on a loan-level to address granularity in credit risk of loan.

Valuing whole-loan pools is difficult without loan-level technology. The best results are achieved using a big-data approach to generate cashflows on individual loans. PeerIQ is a leader in whole loan valuations. Over the next few weeks we will delve into our valuation methodology in detail.

Reach out to learn how PeerIQ can help you with valuations

 

PeerIQ Mentions:

 

Conferences:

 

Industry Update:

Lighter Fare: