Defaults, bad debt, debt sale

Annualized Default Rate

wiseclerk

I just watched the recorded webcast. It’s great that Lending Club uses these to communicate to the users. However I found the way some information were presented to the lenders to be controversial. About 11 minutes into the presentation the company advertises the Annualized Default Rate of 2.36%.  Looking at the slide at 0:13:29 the company states “Less than Three Loans out of the 100 Default”. Is that right – does the percentage of Annualized Default Rate figure match the percentage of loans that default?

This does not match Lendingclub’s own definition of Annualized Default Rate, which is:

Annualized Default Rate is calculated by dividing the total amount of loans in default by the total amount of loans issued for more than 120 days, divided by the number of months loans in default have been outstanding and multiplied by twelve. The loans issued for less than 120 days are excluded from the calculation because loans are unlikely to default during the first 120 days.

I’ll create the following example to illustrate what Annualized Default Rate means to lenders. Imagine a bad-lucked lender that loaned 10 loans with 100 US$ each 12 months ago. First all went well, but after 10 months suddenly 5 of his borrowers failed to pay and defaulted. Colloquially that lender might swear: “That sucks, 50% of my loans defaulted”

Under the formula this gives us an annualized default rate of 8.3%. That sounds much better, doesn’t it? The important difference is that the annualized default rate figure is just a snapshot taken right now. It will rise over the time until the loans mature (if the lender does not invest in new loans). So after 36 months it will be much higher while the figure “50% of my loans defaulted” will not have changed after 36 months (if the other 5 loans continue to be paid on time).
You may want to ask, if the figure could fall instead of rise? No, for a given portfolio the annualized default rate can only go up over time – no loan can return form a default but addituionally further loans could default.

So what does that mean?

First: An annualized default rate of 2,36% does not match the message “Less than Three Loans out of the 100 Default”.
Second: Most of Lending Club’s loans are very young and the overall loan volume is growing. So even if – due to growth – the annualized default rate stays at 2,36% overall, it will rise higher for given loan portfolios orginated in the past. (Compare: ‘Lending Club Default Rates Much Higher than Initially Expected?‘).

Note that the same effects impact the Net Annualized Return rate.

Little Progress in MYC4’s Recovery Attempts

wiseclerk

As reported in the past MYC4.com has serious operational problems making it an investment with negative ROI for the vast majority of lenders. MYC4 has taken measures to recover as much of the outstanding loan amounts as possible, but progress is very slow.

This is a quick update on the situation

Kenya / Provider Ebony:
The receivership has been in place for two months now, but has recovered only a small amount. The court case against Ebony Capital Ltd. is ongoing still awaiting a ruling. (see details)

Ivory Coast / Former providers Ivoire Credit and Notre Nation
The responsibility for collecting these loans has been turned over to TRIUM International in September 2009. In the 5 months since then TRIUM International collected 17,848 Euro. TRIUM has asked to be relieved of the contract as soon as possible (see details)

Senegal / Provider Birima
Repayments have been delayed. Birima cites technical problems and a bad economic situation in Senegal.

Uganda / Provider FED/CMC
FED seems to have the worst status. MYC4 reports that collections nearly stopped due to a lack of staff and  working capital. Borrowers are said towithhold repayments in speculation on a collapse of FED/CMC.
MYC4 has defined 10 action steps for March and April. (see details)

Continue reading →

Lending Club Default Rates Higher than Initially Expected?

wiseclerk 3 Comments

Back in January I received an email from a Lendingclub employee in reaction to this article, where I wrote:

“… Several .. p2p lending services show clear signs that default levels will (or have) surpassed the initially published percentages of defaults to be expected based on external data. … The one exception from the rule is Zopa UK, which successfully manages to keep defaults low…”.

The email questioned why Lending Club was not mentioned along Zopa for keeping defaults low and invited me to discuss this. On Jan. 21st I replied with the following (based on numbers which I compiled from Lendingclubstats.com – these will have changed slightly since then by now):

As sample let’s look at the loans Lending Club issued in Dec. 2007. Total loan amount is 1,322,850 US$.

The status of these is:
a) Current 823,800 (62,3%)
b) Fully Paid 168,150 (12,7%)
c) Late 82,500 (6,2%)
d) Defaulted 248,400 (18,8%)

These loans were approx. 2 years old (in January) and will run about 1 year more.

Is it a fair assumption that in Jan 2010 22% (or more) of the loans issued will have defaulted? I know I did not take the final step to split these numbers by credit grade, but if I would have done that, are you arguing that the default levels are low (or at least lower than the scoring predicted in Dec 2007)? If Dec. 2007 is for some reason a bad performing month, feel free to do the above with any other month from 2007 for the discussion and we continue with these numbers.

Though I was promised a detailed answer and I did follow-up several times, so far there has been no reply. I am not saying that Lending Club defaults are too high for lenders to make a profit. My points are:

  • Default levels at Lending Club are likely higher than initially expected
  • The published default rates on Lending Club and other p2p lending platforms are often averages in relation to all running loans (including recently funded ones). This figure is skewed, if the service is growing fast and lenders might misinterpret it. A better  evaluation is based on taking a sample of older loans (e.g. based on one month of origination)
  • ROIs for Lending Club lenders will be, once their investments mature, likely lower than the average shown at the moment at the Lending Club statistic page.

For Debate: Can Data from Social Networks be Used to Reduce Risks in P2P Lending?

wiseclerk 7 Comments

P2P Lending is mostly anonymous and loans are unsecured. To make the risks of lending to a stranger acceptable for lenders, p2p lending services had to provide models for the lenders to judge the dimension of the risk of not getting paid back.

The initial estimation of the risk-level could not come from the platform itself as it had no track record and could not build a model that “calculated” the level of risk involved for the lender. The consistent consequence was that nearly all p2p lenders relied on established third party providers for credit history data and credit scores. Prosper for example showed Experian data on default levels to be expected depending on credit grade.

Over the time it became obvious that the actual default levels at Prosper were much higher than the expected default levels based on Experian data. We don’t actually need to argue here what led to this (be it financial development of the economy, be it that p2p lending attracted bad risks, be it a poor validation process), but the result was that since defaults were much higher than expected, lender ROIs were much lower than expected at the time of the investment.

And this is not Prosper specific. Several other p2p lending services show clear signs that default levels will (or have) surpassed the initially published percentages of defaults to be expected based on external data.

Boober failed due to default levels, on Smava levels are higher than the Schufa percentages fore-casted, same is likely for Auxmoney defaults which will be higher then Schufa and Arvato Infoscore data suggested. The one exception from the rule is Zopa UK, which successfully manages to keep defaults low, as CEO Giles Andrews rightly points out.

Continue reading →

Prosper’s Legal Collection Test Result Fail Expectations

wiseclerk

Prosper has published a review of the results of a legal collection test. In November 2007, Prosper had selected 74 loans with an outstanding principal balance of approx. 704,000 US$ to conduct a test for a legal collection strategy instead of including them in a debt sale (which at that time was the usual Prosper procedure for bad debt).

The cases were handed over to the law firm Hunt & Henriques.

Since then there was none or little official communication about the progress. Relying on other sources, P2P-Banking.com reported last year that several of lawsuits in these cases were lost.

The new blog post by Prosper describes in detail which steps were undertaken and what results the measures yielded. The only step that can be counted as somewhat successful was the pre-legal phase of letters threatening lawsuits which recovered about 40,000 US$ payments. 66 accounts then went into the legal process.

Surprisingly 16 cases (24%) had to be closed because the debtor moved out of state (3) or Prosper was unable to obtain service.
On a sidenote: Interested parties have raised the questions why Prosper did not apply to the court to allow service by publication, which seem to legal and often used in California as P2P-Banking.com was told. In this case, after other measures failed the plaintiff runs an classified ad in a newspaper. It does not matter if the defendant actually sees this newspaper ad.

The remaining 50 cases further dwindled when Prosper deducted cases with bankruptcies and lowered credit scores which it deemed not worthwhile. Continue reading →