It's no secret that consumer credit that is originated over the internet is the asset class du jour
Stories published by Bloomberg and the Wall Street Journal illustrate how "peer-to-peer" loans are being embraced by institutional (large and small) money managers alike. Banks are even starting to provide investors with leverage to increase their exposure to consumer credit.
I can personally attest to how the involvement of institutional investors has affected the p2p lending landscape. Back in March of this year, I blogged about a couple of my investing strategies. In particular, I had set up some filters on LendingClub's website (which I have since changed) and could log-on any time during the day and peruse a dozen or so loans that met these criteria. A few months ago, I started noticing that I needed to log-on at the specific times that LC posts new loans (i.e. 6 a.m., 10 a.m., 2 p.m., and 6 p.m. Pacific time) in order to find loans - at least then, I still had time to look through the loans to decide which ones I wanted to invest in. Fast forward to today - by time I run my filter and wait for the results, I am lucky if there are 2 loans available to be funded. And if I take any time to review the loans, it will be too late.
There are many good suggestions on how to increase availability of loans both for institutional investors and for individual investors - there is a long thread on Lend Academy's forum discussing this very issue
But one suggestion that I have not seen yet, but which I think is worth exploring, is to do what 'wall street' does when it wants to provide exposure to a asset class - Use a DERIVATIVE
The most obvious form of derivative that can be used to replicate the returns on consumer loans is the good-old-fashioned credit default swap. For those of you unfamiliar with a CDS contract, there are plenty of good primers online. But in short, a CDS contract requires:
1. Protection Buyer (i.e. person who is short the credit risk)
2. Protection Seller (i.e. person who is long the credit risk)
3. Reference Entity (i.e the actual borrower)
Under the terms of the CDS contract, the Protection Buyer agrees to pay a periodic payment to the Protection Seller (the Premium) based on the Notional Amount of the CDS contract (i.e. 10% p.a. on $100). If the Reference Entity defaults (i.e. the borrower fails to repay a loan), the Protection Seller is required to pay to the Protection Buyer the notional amount of the contract less any amounts recovered from the actual borrower.
Some of you reading this will say, "wait...are CDS contracts weapons of mass financial destruction?"
Undoubtedly, CDS contracts facilitated excessive risk taking by certain institutions prior to the 'great recession'. But arguing that CDS contracts are inherently harmful and should be banned is the same as arguing that "guns kill people so they should all be banned" (actually, i personally think guns kill people and they should be banned but that's for a different post.....)
Having spent many years both in the legal profession covering derivatives transactions as well as a few years on a CDS trading desk, I've had the opportunity to see many of the issues that have arisen from the use of CDS contracts so I am confident that CDS contracts on consumer loans can develop in a way which eliminates many of the issues experienced in the traditional CDS market.
How a CDS contract on a consumer loan could work
1. Protection Buyer and Protection Seller agree on a Reference Entity (e.g. LendingClub ID 3496360), term (e.g. 2 years), Premium (i.e. 15% p.a. payable monthly) and Notional Amount (e.g. $100)
2. Outside a default by the Reference Entity, the major risk associated with CDS contracts is a default by the Protection Seller. This is essentially what happened to AIG - AIG was a protection seller on very large notional amount of CDS contracts. When the underlying reference entities (which happened to be mainly residential mortgage ABS) started to default, people worried that AIG would not be able to actually pay the protection buyers.
To address this risk, the Protection Seller on a consumer loan CDS contract will be required to fund the entire Notional Amount of the CDS contract into a trust account. Requiring the Protection Seller to fund the entire notional amount removes the inherent leverage found in a traditional CDS contract and also eliminates the risk of a Protection Seller default/bankruptcy. This should not however deter the current p2p loan investors from becoming Protection Sellers since they are required to fully fund their current investments.
3. Prior to a Reference Entity default, the Protection Seller will receive the Premium on a monthly basis. However, as opposed to tradition CDS, where the Premium is simply paid by the Protection Buyer on the Notional Amount of the contract, the Premium for a consumer credit CDS contract will be a dollar amount equal to the monthly payment amount that would be payable as if the CDS contract is a fully amortizing loan. The Protection Buyer will pay the 'interest' component of each Premium payment and the principal component of each Premium payment will be paid by a repayment from the from the trust account to the Protection Seller.
4. If the Reference Entity defaults, all amounts left over in the trust account are paid over to the Protection Buyer. I understand that this provision assumes a recovery of 0 but given the nature of the credit, I think this assumption is warranted and can be priced into the Premium.
5. Another issue with traditional CDS contracts is defining what is a Reference Entity default (or in CDS parlance, a "Credit Event"). Without going into any detail, the International Swaps and Derivatives Association has published its definition of what constitutes a "Credit Event" and there are committees established to determine whether or not a Credit Event has occurred with respect to a particular Reference Entity. If you really want to find out more, click here.
However, if the Reference Entity is simply a LendingClub or Prosper note, the Protection Buyer and Protection Seller can simply agree that once the Reference Entity is "Charged Off" by Lending Club or Prosper, a 'Credit Event' will have occurred.
6. One issue that the Protection Seller and Protection Buyer will need to agree on is what happens if the Reference Entity pays-off the loan prior to the end of the CDS contract. Fortunately, this issue was addressed by the short-lived 'loan-only' CDS market. Again without going into details, I suggest that the consumer credit CDS market follow the European model and simply cancel the CDS contract if the loan is prepaid. The Protection Seller is refunded all the money left in the trust account and the Protection Seller stops paying the premium.
7. Finally, the last issue that needs to be addressed is what happens if the Protection Buyer defaults (i.e. stops making the Premium payments). In traditional CDS contracts, what would happen is the Protection Seller would mark-to-market the CDS contract and depending on the market value of the CDS contract, either the Protection Buyer or Protection Seller would be required to make a payment to the other party. Since it would be very difficult to mark-to-market a consumer loan CDS contract, I suggest that a default by the Protection Buyer simply result in the cancellation of the contract.
So now that I've provided a framework for how a consumer credit CDS contract could work, let me know what you think.
In particular, who do you think would be natural protection buyers? I think credit card companies, local banks and credit unions may be good candidates to use consumer credit CDS as a hedge. There may also be institutions out there that think consumer credit is a bad risk and want to short this burgeoning market.
Coming up next time......
I recognize CDS contracts on individual LC or Propser notes is never going to take off given the idiosyncratic nature of each particular credit. However, what could work is creating an index of 10000 or so LC or Prosper notes. I think this product would be a great proxy for consumer credit risk and lead to an active 2-way market so stay tuned....
Table of Contents
Regulatory Issues. 4
A.... Securities Laws... 4
1...... Securities Act..... 4
2...... Blue Sky Laws..... 7
3...... Securities Exchange Act..... 8
4...... Investment Company Act..... 9
5...... Investment Advisers Act..... 10
6...... Risk Retention Requirements..... 12
B.... Lending Laws and Lender Registration/Licensing... 14
1...... Usury Laws..... 14
2...... Other Financial Institution Regulations..... 15
3...... Issues Related to Third-Party Use of Bank Charters..... 16
4...... State Licensing Requirements..... 17
C.... Borrower Protection Laws... 17
1...... Truth in Lending Act..... 18
2...... FTC Act and UDAP Laws..... 18
3...... Fair Lending Laws..... 19
4...... Debt Collection Practices..... 20
5...... Privacy Laws..... 21
6...... Electronic Commerce Laws..... 22
7...... The Dodd-Frank Act and Consumer Protection..... 22
D.... Bankruptcy Considerations... 23
More Information. 27
i also don't invest in "G" rated notes for the simple reason that, according to LC, the historical returns on "G" rated notes is lower than "F" rated notes. assuming that "G" rated notes are in fact riskier than "F" rated notes, investing in "G" rated notes would violate every risk/reward principal such as investing on the efficient frontier, etc.1
i don't suspect that the above referenced filter differ too much from what others would consider fairly common sense filters. there are also a number of other filters that people may employ but for one reason or another, i choose not to
however, i think the following filter does differ from what many other investors think and also greatly reduces the availability of loans that i will investment in. the filter is "Loan Purpose" - according to LC (as of today), 78.15% of borrowers apply for a LC loan with the express intent of paying off credit cards or consolidating their loans. i simply don't invest in those loans.
this may sound crass but in my experience, human behavior does not easily change. yes, there will be individuals who recognize that their previous spending is unsustainable and are now truly committed to reducing their debt burden. but since p2p lending should not be thought of as loans to individuals but rather as an investment in an asset class, i am of the view that most people who have incurred significant levels of debt and now require more debt to retire old debt will have difficulty changing the behavior which has resulted in their current predicament.
so with respect to "Loan Purpose", i screen for borrowers who at least purport to be using the loan proceeds for "home improvement projects", "business loan", "major purchase" and a few others.2 certainly some will say that my pool of borrowers is adversely selected meaning that they are borrowers who could not obtain credit through 'traditional means'. on the other hand, i view these borrowers as people who (A) can't be bothered with the hassle of obtaining an unsecured bank loan (if that's even possible these days) and (B) could finance their purchases or activities through credit cards but are savvy enough to explore alternative/cheaper financing solutions.
once i apply the filters mentioned above, i generally fund all available loans at the minimum amount ($25). because my filter does eliminate a vast majority of available loans, i find that i am only able to fund about 10 loans a day and of the loans i choose to fund, i find that about 50% actually end up being issued.3 as a result, it has taken a number of months to build my portfolio (and i continue to add to the portfolio on a daily basis). currently, my portfolio has a weighted average interest rate of 13.64% comprised of the loans with the grades shown below. the portfolio is still relatively young but so far, i have not had a loan enter a 'late period' and continue to re-invest payments as i receive them. unfortunately i haven't taken the time to actually calculate my returns on this portfolio. the reason why the 'net annualized return' of 9.12% quoted by LC with respect to my account is not accurate is that i have funded other notes on LC utilizing other strategies include one that turned out pretty badly. when i get around to calculating the actual return for this strategy, i'll be sure to let you know.
Composition of my LC Notes using the strategy described above
So Step 1 of my analysis is to try and realistically answer questions 1, 3 and 4 (question 2 will be addressed later) - in order to answer these questions, i try to (a) determine the year in which i think an exit is realistic (either by a sale of the company or a subsequent round of financing into which i can sell), (b) estimate the earnings of the company in that year and (c) determine an appropriate P/E ratio for the company1
i generally don't have much basis for making the assessment in sub-step (a) other than what information the company provides. absent any specific guidance from the company, i typically model out a 4 year investment horizon. i feel that 4 years is an appropriate amount time for a company that is seeking an angel investment to either sell itself, raise another round of financing (into
which i would be able to sell my shares) or fail
sub-step (b) (estimating earnings) is in and of itself a challenge which i'll try to address in a future post - again, i rely on the company's projections and also model my own projects using what i consider reasonable estimates for growth rates (both of revenue and expenses)
finally, in order to complete sub-step (c), i simply use publicly available resources to find an appropriate ratio based on the industry that the company is situated2
after determining the exit strategy and 'terminal value' (i.e. how much the company will be worth at the time i hope to exit), Step 2 of my analysis is to determine the IRR on the investment (based on the price of the investment and assuming results of Step 1 are accurate)
generally speaking, angel investments are either in the form of equity or convertible debt. in the case of equity, it is straight forward to determine the cost of the investment. however, in the case of convertible debt, a few additional assumptions are necessary which i will address when reviewing an actual investment
the purpose of Step 2 is simply to check if the investment will generate a return above a certain threshold - in my case, i've set the threshold at Step 2 of 60% p.a. (or roughly a 6.5x cash-on-cash return assuming an exit in 4 years)
finally, in Step 3, i apply a 'probability matrix' to determine the range of possible returns - this step involves coming up with 48 different return scenarios by varying (a) the dilutive effect of subsequent rounds of financing, (b) the time frame before i am able to exit and (c) the value of the company at the time of exit. i then assign a probability to each scenario and determine if the probability weighted return still satisfies my return objective (15% p.a.) since i am coming up with 48 scenarios in which the company actually survives, this leaves 52 scenarios in which the company fails (and my investment is a complete write off). based on the literature, a 52% failure rate by start-ups is probably too low if one looks at the entire universe of startups. however, because i am investing through crowd-funding portals that (purport to) vet the companies i am comfortable (at least for now) with this assumption. with some time, there should be more statistics available that demonstrate the success/failure of companies that raise funds through various crowd funding platforms so i'll be able to tune this assumption accordingly
the following are my typical scenario adjustments - however, i'll tailor the adjustments depending on how much volatility i believe there is in the potential outcomes:
of course, Step 3 only works if i can appropriately assign a probability to each scenario. given how little control i will actually have over the company, i currently assume that each outcome is equally possible (i.e. the is a 52% chance that the company will fail and a 1% chance that one of the other scenarios will occur) - i know this is a pretty simplistic approach so any input you may have will be appreciated
Application of General Framework to Company 1now that i've provided some background information on my frame work for financial analysis, i'll try to show how this works in practice
angel.cowhile all of these portals are slightly different in the way they operate, the basic premise is (1) companies post information about themselves; (2) investors browse the companies; (3) investors invest in companies they like (typically in the form of equity or convertible debt) and (4) everyone profits4
[please add to the list in the comments section]
http://www.angelresourceinstitute.org/5. due to confidentiality agreements as securities laws, i'm not going to provide exact details but will provide (hopefully) enough information for you to understand the company and help with due diligence and valuation
|Probability of Default|
|Prosper Grade||Average FICO||Lending Club Grade||Average FICO|
As always, please let me know if you have any questions or comments
When deciding how to best finance the necessary upgrades, City officials contacted their financial adviser who in turn, submitted a RFP to a number of financial institutions. After a number of steak dinners and Broadway shows, the City of Detroit chose XYZ Bank (the “Bank”) to underwrite the bond issuance.
In order to protect the City against projected increases in borrowing costs, the Bank and the City entered a 30 year vanilla interest rate swap (a.k.a a derivative) where the City would pay a fixed rate of interest and receive a floating rate indexed to 3-month LIBOR. And because the City was rated AAA and the Bank wanted to maintain its client relationship, the Bank agreed that the City would not be required to post collateral based on the mark-to-market of the swap. Prudently, the Bank hedged the interest rate risk associated with the swap by entering into offsetting swaps with other financial institutions.
Fast forward to 2011 – notwithstanding the Lions making the playoffs and the success of Eminem’s Recovery album, unemployment rates in the City are significantly above the national average, two of the three major automobile manufacturers have required substantial government assistance to remain afloat and the City’s tax base has been deeply eroded. Auctions for the bonds fail and the City is now required to pay 12% interest on the bonds in order to prevent a default. Nationally and around the world, central banks have aggressively fought to prevent a global recession by keeping borrowing costs at all time lows. As a result, the value of the swap between the City and the Bank is significantly (for the purposes of this example, let’s say its $50 mm) in favor of the Bank.
It is a popular misconception that the Bank has now made $50mm and will pay that money out to its employees in the form of bonuses. In reality, the Bank also feels substantial pain:
So what happened? Many believe that the Bank must have hoodwinked the City in order to make a quick profit. Some would go as far as accusing the Bank of committing fraud. Banker’s and their lawyers argue that the City was appraised of the risks and independently decided on a course of action. Regulators and lawmakers believe that more rules would have prevented this story from playing out.
What happens next?
Again, the following is hypothetical but I can imagine a scenario in which:
Notwithstanding the fact that the swap itself may be $50 mm in favor of the Bank, the Bank also had significant exposure or costs associated with swap.
1. It needs to fund the margin requirements on its own interest hedges
2. It needs to write down (or at least reserve) some or all of the $50 mm due to the credit riskiness of the City (CVA charge)
3. It is (or will be) required to hold a significant amount of additional capital against the position when certain provisions of Basel III are implemented. The associated lower return on equity will put downward pressure on the Bank's stock price.
In any event, terminating an uncollateralized derivative is always beneficial for a swap dealer. Although I am not suggesting that the swap dealer must disclose its funding levels and open positions, etc., I do recommend that municipalities, sovereigns or any other entities that are party to an un-collateralized derivatives fully review and understand the transaction before agreeing to pay any termination fee.
Please note that this (and all) blog post is for informational purposes only. It
does not constitute legal advice and is not intended to create an
attorney-client relationship. Readers should not act upon any
information presented on this blog without seeking their own professional legal
counsel. Click here for additional disclaimers and information.