Concluding remarks
We identify the trend in portfolio performance by applying only three consistent scenarios scaled up from canonical to extreme conditions. We showed that, under canonical conditions, a diversified portfolio of marketplace loans, based on our sample portfolios, stands to earn roughly 23.9 percent over the term of the investment (60 months), or 6.63 percent per annum.
However, under mild stress conditions, the total portfolio returns decline to roughly 4 percent or 1.43 percent per annum. Under stress, total returns turn negative to -14.71 percent or -4 percent per annum. This is by no means horrible, but it is also a far cry from the sure-fire value proposition that entices retail investors into marketplace lending. Notably, we performed more simulations using different portfolio sizes (e.g., from 100 to 2,000 contracts per portfolio). The results, however, under the same scenarios are very similar to the ones presented in this chapter. Thereader is welcome to apply the analysis described in this chapter, considering the predefined or their own sample data and model parameters, by using the software accompanying this textbook.Instead of median expected returns under ideal conditions, investors must be aware of the real possibility of their investments turning negative, with no way to cash out before the term of the illiquid investment has run its course. Holding on to an investment with a negative return over five years might do much damage to the public perception of the sector. Marketplace lending platforms should do everything they can to avoid lenders experiencing returns such as in Scenario A and Scenario B.
Nevertheless, platform operators argue that it is the responsibility of the lender to analyze the risk-and-return characteristics of loans on their platforms. They are only partially right. Platforms oversimplify the asset class by showing only the expected returns under fair weather conditions and omitting the potential losses that might occur under stress.
It would be good practice to improve transparency in this regard. It should be in the interest of platforms to offer suggestions to lenders that help them avoid losses even in difficult markets. Platforms have no empirical experience of a full credit cycle, so they need to use models to simulate potential outcomes. We hope they run similar analyses like we have done in this chapter. Unless they do so, platforms risk saddling their investors with losses they didn't see coming. This may drastically decrease the appetite of the investing public in the asset class. Also, if the return of a loan book of a platform is negative, how exactly will marketplace lending platforms uphold their high valuations? Lending platforms should aim to offer their investors the ability to structure the best portfolios possible that will be profitable under various conditions.Institutional investors and other professional investors that allocate capital to marketplace loans have access to sophisticated analytics of third parties that simulate scenarios like the ones we have investigated in this chapter. However, relying on institutional capital only can be a risky strategy for lending platforms. Because their funding sources closely correlate, platforms risk a freeze in funding liquidity under stress conditions. This could be the death knell for many players in the sector. With this in mind, what should platforms and investors do? They must identify the following key issues in their portfolios.
■ What are the risks that portfolios are exposed to?
■ Which are the contracts that are most sensitive to these risks?
■ What is the impact of stress on performance and losses?
After clarifying these questions, platforms have the following options:
a. Restructure portfolios by employing contracts that are least sensitive to risk and providing best performance. Investors can do this over time by rolling over their portfolios or by selling the underperforming contracts and replacing them with new ones.
b. Collateralizing and/or mitigating risk exposures. The former needs additional effort of identification, measurement and monitoring, whereas the latter increases the cost of loans.
c. Do nothing and hope that investors’ assumptions will come true and no additional uncertainty creeps up.
Most practitioners apply a combination of the above, and investors in marketplace lending should do the same. With the appropriate tools, platforms and their investors should be in a strong position to bolster portfolios against unnecessary volatility.
NOTES
1. Provided by Lending Club and referring to the year 2014, available on their website; Lending Club Statistics (2015a), Download Loan Data, https://www.lendingclub.com/info/download-data.action, date accessed April 20, 2015.
2. Based on Lending Club data, the interest rate varies from 6 percent to 26 percent, Lending Club Statistics (2015a), Download Loan Data, https://www.lendingclub.com/info/download-data.action, date accessed April 20, 2015.
3. Variable interest rates tend to adjust the contract’s value and liquidity according to market conditions.
4. 1% to 3% for a duration of 30 to 70 months.
5. In our case study we interpolate the spreads based on interest rates defined in Lending Club Statistics (2015a), LoanStats3c.xlsx. In this case study they range from 1% to 35% for A1 to G5 ratings per annum
6. The approach of conditional default probabilities based on a hazard rate linked to credit spreads and recovery rates is also the basis of Basel III credit and counterparty risk and valuation adjustment of credit portfolios (see BIS paper bis.org/publ/bcbs189.pdf).
7. The approach of conditional default probabilities based on hazard rates linked to credit spreads and recovery rates is also the basis of Basel III credit and counterparty risk and valuation adjustment of credit portfolios (see BIS paper bis.org/publ/bcbs189.pdf).
8. Using normal annual time interval.
9. E.g., BIS in Basel III regulatory papers.
10. Lending Club (2015a), “Automated Investing Allocation Summary” under account, https://www.lendingclub.com/portfolio/primeReport.action, date accessed 21 May 2015.
11. A detailed illustration of all sub-rating distributions among 1000 contracts can be seen in the file “Loans.xlsx” in the software model accompanying this textbook.
12. This impact of counterparty credit status to credit exposer, named specific wrong way risk, is discussed in Chapter 11 (Credit Enhancements).
13. General wrong way risk occurs when market performance impacts negatively the exposure.
14. 30 to 60 months.
15. In this exercise we use the interest income as recovery.
16. Lending Club (2015a) “Rates and fees,” https://www.lendingclub.com/public/borrower-rates-and- fees.action, date accessed 21 May 2015.
17. Detailed data referring to shock factors are provided in Annex B, available on the website, together with the model.
18. Detailed data referring to applied default PITs are provided in Annex B, available on the website, together with the model.
19. Lending Club (2015a) https://www.lendingclub.com/info/demand-and-credit-profile.action.
20. http://www.wsj.com/articles/lenders-step-up-financing-to-subprime-borrowers-1424296649.
21. http://www.biz2credit.com/research-reports/analysis-bank-failures-2009-2014-identifies-credit- desert.
22. Dugan, Ianthe Jeanne and Demos, Telis (2014) “New Lenders Spring Up to Cater to Subprime Sector,” Wall Street Journal, 5 March 2014, http://www.wsj.com/articles/SB1000142405270230473 2804579421653206982012, date accessed 18 July 2015.