Friday, December 23, 2011

A Revelation On My Grace Period Epidemic

So for those just joining us, my Lending Club experiment was going well since August, in fact, very well. So well, that I shifted all my fun investment money over to Lending Club and have regular deposits. Then in November I started getting Late Notes. I figured it would happen sooner or later. Then I started getting a lot of Late Notes. I also noticed that Late Notes don't sell very well. So I started discounting Grace Period Notes. If it doesn't go into Grace Period, it won't be late.

I noticed that I had a lot of Grace Period Notes. Since I wasn't even keeping track of Grace Period before, I just figured it was that way all along and was common. I attributed the Late Note epidemic to the season (and possibly the economy). Now that I am discounting a lot of Grace Period Notes, I feel it is affecting my return. However, I think I could have avoided most of this mess if I just understood what Never Late meant.

The way I look for Notes is to use all Percentages and select Never Late and Now Current (disable the other Now Late checkboxes). Then I sort by Yield To Maturity. My assumption was that every Note in the list had always paid on time (never in Grace Period).

Today I was using my Cash to buy up Notes when I noticed a steeply discounted Note with the highest Yield To Maturity. When I investigated, I found the following:
In the Browse Notes list it showed it was Current. However, they were still trying to process the payment. It was steeply discounted because someone was watching their portfolio closer than just the status reported by Lending Club. I would consider this Grace Period, but at least now I know.

So now I will screen each Note before purchase (clicking on the Current or Issued link) to see if it really is Never Late. I expect that would cut down on my Grace Period epidemic.

Monday, December 19, 2011

Five Months and The Good Times Roll (For Those Who Ran Off With My Money)

(See other Monthly Status Updates)

So after five months I've had a great run, followed by an avalanche of Late notes. I think I've had a total of 13 notes that have gone Late. One of them I've been able to sell off.

Every one of the Notes that have gone Late have made payments, but none of them have made payments to me. They were never late and were current when I bought the Notes (assuming that is what those checkboxes mean when Browsing Notes).

When they give my return, they obviously don't take into account that the Late notes are as good as gone. The chart above shows that of the Notes that go past 31 days, you have basically a 50/50 chance of getting any money back (they don't specify how many were fully recovered versus partially recovered, and how much that partially is). The site reports my return as over 21%. They no longer show your percentile. When they did show it, I had dropped solidly into the 99% (before I was waffling between being in the top 99% and 100%). My own internal reporting needs to change. I've been calculating my account value as the sum of what I paid for all my notes. What I need to use is the lesser of what I paid or what I am asking. I believe that will more closely represent my account value.

Of course, it may be more representative to consider Late Notes to be $0. I don't know what the conversion rate from Grace Period to Late is, but from 16-30 Days Late to 30-120 Days Late has been 100% for me. I've been able to stem the tide of Late Notes by being more aggressive at selling Grace Period Notes. I've still had a few slip into Late status. 

I've been selling Grace Period Notes at 80% of P+I (rounded down to the nearest dollar), and dropping the price by $0.10/day. Some of these Grace Period Notes have received payment and then sold at this discount. At first I was disappointed, but then I realized I didn't want the notes where the borrower was willing to go into Grace Period.

Late 16-30 Days I sell at 75% of P+I (rounded down to the nearest dollar), and dropping the price by $0.10/day. Late 31-120 Days I sell at 50% of P+I (rounded down to the nearest dollar), and dropping the price by $0.10/day. Not that discounting Late Notes seems to work. I have only sold one Late Note, and none have returned from the grave. I may consider being more aggressive at selling Grace Period Notes just to prevent Lates, since they seem to be the land of no return.

Most of my Notes (almost 2/3) are still F Grade. If I read the numbers correctly, 76% of the F Grade Notes are still Active (Current or Late). Of the Active F Grade Notes, 95% of them are Current. 25% of my Late Notes are F Grade while the compose 63% of all my notes. It appears that F Grade Notes are a pretty sweat spot.

E Grade Notes have roughly the same Current, Late and Default rates. They are also about 25% of my Late Notes. However, they only comprise 6% of my portfolio. So apparently I've been really unlucky with E Grade Notes.

Half of my Late Notes are G Grade. Given that about 1/3 of my portfolio is G Grade and that the stats for G Grade notes is a bit worse (91% of Active Notes are Current as opposed to 95% for E and F) that is not surprising. About 8-9% of E and F Grade Notes are in Default. For G Grade Notes, it's 14%.

You would think that due to the very small place that E Grade Notes hold in my portfolio (and that they are statistically equivalent to F Grade Notes) that they would not hold such a prominent spot in my Late Notes.

It is also interesting to see what Lending Club is doing to attempt to reclaim my money from the Late Notes. Below is a chart of the latest action Lending Club has taken on the Late Notes.

Months OldPayments MadeLatest Action
32Sent email to borrower
47PAYMENT Failed
52Borrower contacted Lending Club
53Sent email to borrower (three notes for this loan)
84Borrower located (skip trace)
85Borrower provided Bankruptcy counsel information
96Borrower filed for Chapter 7 Bankruptcy
96Collections Agency attempted to contact borrower
128Drafting lawsuit
1916PAYMENT Failed (previously: Borrower provided Bankruptcy counsel information)

So now after my big Late Note hit, I'm making some changes. First, Grace Period Notes seem to sell. I think I need to discount them slightly more than I have been. I'll be discounting them to 75% of value instead of 80%. Late Notes don't seem to sell at all (especially 31-120 Days Late). I'm also adjusting my Risk Factor for Note Grade. I used to just convert the letter to a number and the trailing number to the fraction (ie A1 = 1.1, F4 = 6.4). Now I will be squaring that (ie A1 = 1.1^2 = 1.21, F4 = 6.4^2 = 40.96). This should cause my tool to more steeply discount G Grade Notes, as they will be rated 36% more risky instead of 17% more risky. This may also weight my portfolio more towards E Grade Notes. Given the lower return and about equal risk to F Grade Notes, I'm not sure how I feel about that (not to mention my propensity to have Late E Notes).

Also, of note may be the states my Late Notes are from: SC, AZ (2 Notes), NY, VA (3 Notes, same loan), CA, OH (2 Notes), OR, MN. I was already weighting the risk of Notes based on state, thanks to Nickel Steamroller. Currently I weight riskier than average the following states (riskiest to less risky): CA, FL, NE, IN, MS, TN, IA, ID, MT, UT. I also weight the following states less risky than average (least risky first): WY, ME, OK, LA, CT, KY, NC, WV, KS, AL. I am adding "riskier than average) to my Late Notes' states (although I already had CA as the riskiest, I considered the others as average risk). Since SC and OH have two Late Notes each, I made them as risky as CA and FL. The other states I put their risk about equivalent to TN or IA.

And yes, I have three Notes from the same Loan and the Loan went into Default. I have identified some holes in my detection of Notes from Loans I already own. This is the risk I was trying to avoid. I think I've worked out a way to avoid the hole that got me three notes from the same loan and I think I discount Notes which I have multiple from the same loan.

If I assume that all my Late Notes are worthless, I think I'm basically back to my original investment, so 0% return. So now that I've learned a lot of lessons, I'm starting from the beginning again, and hopefully will come back with a vengeance!

Monday, December 5, 2011

'Tis The Season, I Hope

 The first three months were great: August, September and October. Since then, I've had 12 loans go late. One was sold, the remaining 11 I still have (not by choice).

My thoughts right now are that as soon as a note goes into Grace Period, it's time to start discounting it. I was just giving a mild discount (Principal + Interest rounded down to the nearest dollar). This led to three more notes going late. I am now going to 80% below P + I for any notes that goes into Grace Period. This is based on the recovery rate from the chart below.

I still discount 16 - 30 day late notes 75% and 31 - 120 day late notes 50%. Grace Period, Late 16 - 30 and Late 31 - 120 I drop in price $0.10/day ($0.60/week). It may be that I've been once bitten, twice shy, but I'll find a happy medium as I experiment with discounting.

I'm hoping that it is the end-of-year season that is causing the larger number of Grace Period and Late notes, or at least due to fluctuations in the economy.

What I'd really like to see is someone blogging about their experience in taking discounted E, F and G Grace Period or Late notes and making some nice returns from those. If everyone followed the advice I post, no one would be there to pick up the notes I don't want anymore.