Lending Club’s portfolio

What is the current state of crowdfunding? Is it evolving and capturing a larger spectrum of the market?

Initially, potential clients looking into crowdfunding were those who might not have the best credit history. Many of them turned to this new market, in the worst-case scenario, because there was no other option. In addition, companies like Lending Club and Prosper needed clients to prove their idea and consolidate their position among investors and creditors. Their model was proven to be effective in due time; Lending Club, for instance, has originated close to 20 billion dollars since creation.
I decided to look into how Lending Club’s portfolio has changed over time. At their site they have information available to everyone,  including analysis over the returns and a description of risk. I have taken a deeper look into how their portfolio has changed overtime, particularly comparing loans originated during their first four years of operations (initil origination) and compare it to what they funded during the first quarter of 2016 (as a sample of most recent origination).

Lending Club has its own algorithm to define the risk each beneficiary has. Measured as grades (A, B, C, D, E, F, G) with A representing the lower risk. Each grade has 5 sub-levels for which we have data, but Lending Club does not present any analysis on the sub-levels; here, I present the analysis looking into the sub-grade.

In terms of data availability we found the first evolution; the variables provided for the different periods is different. Nowadays, more information is available to analyze each borrower. Many key elements such as revolving credit utilizations and current balance on all accounts was not available before. For all new credits these key features are given as well as others, such as previous defaults and type of loans (for all credit bank-type accounts).

But, how has the portfolio changed? To  investigate, I have selected variables that help us describe the borrower risk:

• Annual income (annual_in)
• Average current balance in LC credit (avg_cur_bal)
• Debt to income (DTI)
• The number of inquiries in past 6 months – excluding auto and mortgage inquiries  (inc_lst_6mths)
• Interest rate (int_rate)
• Loan amount borrowed from LC platform (loan_amnt)
• The number of open credit lines in the borrower’s credit file (open_acc)
• Total credit revolving balance (revolving_bal)
• Revolving line utilization rate, or the amount of credit the borrower is using relative to all available revolving credit (revol_util)

In addition I am presenting variables that are interesting, but that no data was available for during the the period 2007-2011.

• Maximum current balance owed on all revolving accounts (max_bal_bc)
• Total current balance of all accounts (tot_cur_bal)
• Total credit balance excluding mortgage (total_bal_ex_mort)
• Total current balance of all installment accounts (total_bal_il)
• Total bankcard high credit/credit limit (total_bc_limit)

To present the data, I have arrange a summary of both vintages (origination group): one for the loans funded between 2007 and 2011 and the other with the loans from the first quarter of 2016.  In the visualization below, the combo that would be most useful now is the second one,  you can select the variable and type of data (bin or dat by sub-grade). 

 

Findings

Income and debt

The revolving line utilization is greater for borrowers of the vintage of the first quarter of 2016 by 6.5%. Meaning that now, people requesting loans from Lending Club use credit more; this is more risk. To confirm this finding, we can look at the DTI which is 16% higher with respect to LC borrowers  from 2007-2011 vintage. Borrowers have more debt in relation to their earnings. But is it more money, or is just more in relation to their income? Upon investigation into this, I discovered that its just the latter.  Looking at the average income in table 2, we can see that in average the 2016 vintage has more income, but this is only because Lending Club now access more people. If we look at the variable (annual_inc_bin), we see that there is no difference in income by strata.  However, the income by risk grade is different  (annual_inc_sub_grade). Borrowers classified as lower risk (A,B, C grades) have higher incomes (newcomers), and those with higher risk have a lower income in the 2016 vintage. This shift in income is evolving favorably for high-quality borrowers, but with negative relation when talking about the higher risk group. A higher risk is not only perceived, but in addition it comes with lower payment capacity. Debt to income DTI is considerably higher in the 2016 vintage for any risk sub-grade.

 

Use of credit

We know that all borrower types have higher debt. How are they accumulating it and from which sources?

The number of inquiries in past 6 months is lower for the 2016 vintage. Since we know that all borrowers have more debt, this may mean that they have easy access to credit. The inquiries in the 2016 vintage is reduced by half compared to the first vintage,  particularly revolving lines (credit cards) for the lower risk type of borrowers. On average,  borrowers of any risk type from the 2016 vintage have opened 3 additional credit lines in comparison to borrowers from the vintage ’07-’11.

 

Cost

The are many external factors used to determine the cost of credit line within Lending Club’s platform. But, perceived risk is one key feature when determining how much a borrower should pay. Of the vintages that were analyzed, the vintage from 2016 has higher interest rates. During 2007 and 2011 the maximum interest rate was 24.6%; in 2016 the maximum was 26.6%. The interest rate distribution of the portfolio is the same in both vintages.

When looking at the interest rate by risk measure (sub_grade), we can see that Lending Club has changed its strategy. In the ’07-’11 vintage, interest rates for higher risk borrowers were lower by more than 25% on average. This is a significant change that denotes the higher risk perceived in borrowers.

 

Conclusion

Analyzing Lending Club’s portfolio, we can see that crowdfunding in loans evolves into an attractive market. It seems to be a trend the sharing economy has proven many benefits. How could it not, seeing as it actually spreads the benefits through to all stakeholders. Particularly in the microfinance industry, there are many competitors as well as both borrowers and investors who are looking forward to it. On the other hand, we can see how people are working around more debt. This is not necessarily a bad thing, but something to be aware of. More debt is more risk to any portfolio, no matter the type of asset involved.

Companies like Lending Club have a great impact on borrower’s lives. They provide cheaper access to credit as well as a place where investors can find a good span of risk to invest. If you have as little as $25, you are qualified to participate. In the process you get more money than what any savings account would give you, and you get to help someone.  I believe crowdfunding is the future of banking. Small investor have access to better returns and borrowers get a fair quote.