Saturday, March 17, 2012

Lending Club Chrome Extension Version 0.2

(See Other Posts about Lending Club Extension)

After using it myself for almost a week, I'm proud to announce version 0.2 is available for install (see Alright, I'm Ready To Share for original announcement).

Install Note Browser Filter 
(click on this link in Google Chrome)


So let's start in with some of the changes. Next to the Out of Range Color boxes there is now a popup that allows you to choose from preselected colors, hide the matching Note or disable that filter altogether.


I've found this handy when I want to temporarily disable a filter and then reenable it, specifically when I have change left over in my account, I don't expect to get 16% Yield to Maturity on $12 Notes.



You'll notice there are a few more links. These each open a section on this page. Opening the Preferences section allows you to set your Short Term Tax Rate (with a link off to a site where you can look it up), your Minimum Preferred Gain and the discount you want to give on Grace Period and Late Notes (in my experience all Late Notes are the same, hard to sell). Why would the extension need to know this information? Glad you asked.

Now when you Sell or Reprice Notes, two new columns are added giving you the Annualized Return and the Markup. As you modify the values, those columns also update with the new values. The Annualized Return is based on the date you bought the Note (more about that later, see Caveat below) the Order Expiration Date (at the top of the table) the Asking Price, the Payments To Date, the Maintenance Fees, the Sell Fee and estimated Taxes (see Can Someone Check My Math? for details on the calculation).


As an example of what happens when you change the values, the snapshot of the table below is identical to the table above except I pressed the Principal + Interest button at the top of the table. Notice all the Markups went to zero and the returns were adjusted accordingly.



Another cool feature is that if any of the Asking Price fields are every empty (deleted) they will be filled in with the value that will give you your Minimum Preferred Gain from the Preferences. For the Sell Notes page, this will automatically fill in every Asking Price (since they start out blank) with a value that will give you your Minimum Preferred Gain (or Discount based on Grace Period or Late of the Note). When Repricing Notes, just delete the value in the Asking Price box and it will fill it back in with the appropriate value. This feature alone saves so much time. When you have 80 Notes you want to put up for sale, determining what you want the sales price to be and then entering it in can be mind numbing.

Caveat

In order for the Annualized Return (or the Asking Price) to be calculated, the extension has to "see" the purchase date on one of FolioFN's web pages. If you visit FolioFN's My Account and the Note shows up there as purchased (this month) then it will have seen the purchase date. However, if you purchased the Note in a previous month, you'll need to go to the bottom of the My Account page and look at the Trading Statement for the month that you purchased the Note and then change the Showing Notes to All.

All you have to do is view these pages after installing the extension. It will remember the purchase dates of all the Notes it sees. What I did to get things kicked off is open each Trading Statement in a new tab and then went to each tab and changed Showing Notes to All. Then I closed all the tabs. Do that once and the extension will remember all the purchase dates.

Feedback and Donate

I've added to other links, Feedback and Donate. The Donate link just gives an Amazon.com affiliate link you can use to show your support for the extension (or alternately PayPal information). I've spent all the time I really need to for my needs on this. I'll be using the donations as a gauge of how much time I'll spend adding requested features and improving the extension.

The Feedback link will send you back to this blog. For suggestions, problems, questions, etc. just leave comments on this blog.

I hope this helps make FolioFN a more usable part of the Lending Club experience (especially for those of us that FolioFN is the main part of Lending Club we can use). Let me know what you think.



Thursday, March 15, 2012

Eight Months on Lending Club

(See other Monthly Status Updates)

Already another month. My NAR keeps dropping like a rock as Lending Club is coming to terms with those Late Notes that will never pay as they move them to Default and Charge Off. I had almost a dozen Notes that went Late. One completely paid the Note off. One came back into payments (and then I was able to sell the Note). I think I sold another couple.

Even though the graph looks depressing, I am optimistic. I have realized what I have done wrong, and how to prevent future Late Notes (which nearly inevitably lead to Defaults and Charge Offs). Since the first wave I have only had one more Note go Late.

As I peruse the other P2P Lending Blogs (see several excellent ones to the right) it seems that 12% NAR is considered pretty good. It's definitely better than than the Stock Market (see Why You Should Invest in the Stock Market) but my target is 16% or higher, mainly to offset the effects of real inflation (see Not All Inflations Are Created Equal). I'm confident I can get back to 16%.

Once I learned the ropes of Lending Club, I wanted to make sure that the time I was spending on investing was not offsetting my gains (unless your time is worthless, you need to consider this). Now that I have the Chrome Browser extension (see Alright, I'm Ready To Share), I feel like I have finally been able to bring down my time spent to a reasonable amount. Before I used this extension, I was only looking at Lending Club 3 days a week to cut down on the time I spent investing. Selling Notes used to be my biggest time consumption, and that was manually entering dozens (even more when I went to 3 days a week from 6) of sell prices from my spreadsheet. Now my extension enters them all for me (coming in version 0.2).

I'm back to 5 to 6 days a week, but it is a 5 minute checkup as opposed to 10 to 15 minutes. And the time spent will not grow dramatically as I scale up my portfolio. I was having a hard time seeing 400 let alone 800 Note portfolio before. Now I can easily see it.

I've now sold 3 times the number of Notes I currently hold (churn). My Notes held average $24.88 and I've held them on average for 2 1/2 months. The Notes I've sold I've held for just over 3 weeks. My median return on sold Notes is 28.8% annualized, or 38.2% days weighted annualized (the longer the Notes been held, the more weight it is given in the average).


The average rate for my Notes held is over 21%. This is a snapshot. I tend to sell off higher Interest Rate Notes, so my held Notes is a little lower, but we'll get to that later.




Almost all my Notes are 60 month Notes. I prefer it this way. I try to buy Notes with a lot of Principal left, which means that the payments are mainly Interest. 60 month Notes have even more Interest since the Principal is spread out over a longer period.



This is a natural side effect of how I choose Notes. I look for Notes with good Yields, and 60 beats out 36 just about every time.





Overall Composition

I always get concerned when I look at this chart. I really don't like holding G Grade Notes, but I have more G Grade than E Grade (I'd rather the reverse). I do like to see that the majority of my Notes are F Grade (in my opinion best Risk/Return balance).

The D and B Grade Notes are the stragglers when I have less than $25 left to invest, then I go find a good Yield Note with less Principal and these can often be lower Grade Notes.

The nice thing about having portfolios is that you can analyze your Note in different segments (see How I Use Lending Club Portfolios). You'll recall I had separated out my fairly newly purchased Notes from those that I had received at least 2 payments from (my Buttercups and Wesleys, respectively). This helps me see how my buy and sell strategy plays out. I buy irrespective of the Note's information just on it's raw Yield or Discount. It could be a Note that has a plunging credit score, lives in Florida, leveraged to the hilt, only on the job for a month, high debt-to-income G Grade Note, and I'd still buy it because it has a good Yield to Maturity.

However, when I sell is when I look at the details of the Notes. That G Grade Note will get sold for near 16% annualized return markup. While the Notes I would rather keep, the F Grade 10+ year employed, home owner, from Oklahoma, get a higher, closer to 32% annualized return markup. So let's see how that affects my portfolios.


Buttercup

The Buttercup portfolio are the Notes I've recently purchased (and haven't gone bad in any way). As you can see, it skews highly towards the G Grade Notes. This is a symptom of buying Notes based solely on Yield To Maturity.

If you compare this graph to the overall composition graph, you'll notice that I sell many of the G Grade Notes before they reach a second payment.










Wesley

My Wesley portfolio are the Notes that I've held for at least two payments and have not started missing payments. This is what I like to see in my portfolio. Overwhelmingly F Grade Notes, followed by E Grade and a few G Grade Notes (and a smackling of smaller Grade Notes).

Notice I still have an over 20% Weighted Average Rate, but I am more heavily invested in F Grade Notes, which have a lower Default Rate than G Grade, and have a higher average Yield (see Low Risk, High Return Lending Club Strategy).



Strategy Change

I have three buying strategies I use. The first is Newly Issued High Yield to Maturity 60 Month Notes (see the bottom of Seven Months In And Things Are Slowing Down). The second is High Yield to Maturity only. The third is Highly Discounted over 16% Yield to Maturity.

I used the second strategy the most heavily for the longest and have been using all three lately. However, until I get my extension to add risk factors to selling prices, I'm probably going to have to focus on the third strategy (Highly Discounted). Many of the seasoned veterans use this method (although not necessarily limiting it to above 16% Yield to Maturity) and have called into question the High Yield to Maturity model. Without my risk based selling prices, I think I'm going to have to agree with them. Plus Now I have a tool that allows my to visually see Discount vs. Yield to Maturity (see Discounted Notes on Lending Club's FolioFN).

Conclusion

So the effects of the Late Notes from the end of last year are finally starting to show in my Net Annualized Return. We'll see when I actually hit bottom on that. As long as I can prevent future Late Notes (which eventually lead to Defaults and Charge Offs), which so far has been pretty successful and will only get better with the Chrome Browser extension, I think I can climb back up to 16%.

I'm excited about the time savings I am getting from using the Chrome Browser extension. It makes the process so much more fun (less tedious) and cuts my time spent "investing" dramatically. It cuts the time down so much, I feel like I still need to do something when I'm done, 5 minutes is not very long.













Tuesday, March 13, 2012

Discounted Notes on Lending Club's FolioFN

On Peter Renton's blog post Changes Today to the Lending Club Trading Platform on SocialLending.net, user @stilltrackin noted that discounted Notes seemed to be harder to find as of February 2012. After I started using my Chrome browser extension I added discounted Notes to my buying strategy. I hadn't been buying discounted Notes for around 6 months and noticed that they seemed to be a bit thin.

When I view Notes, I filter out Notes with principal less than $15 or more than $25, Notes from the same Loan and Notes with less than 16% Yield to Maturity. With those filters, in 1,300 of the most discounted Notes (22 pages at 60 each), I only found 42 that met that criteria and were marked up less than 3%. There seemed to be a pattern so I threw the values into a spreadsheet and came up with this graph, which shows Yield to Maturity on the Y-axis and the discount on the X-axis.
Remember that this is a selection of what I consider premium Notes (see the criteria I mentioned above). It is interesting that there are a handful of Notes right around par and then nothing until around 2%. It is possible that someone is rapidly snatching up everything under 2% markup. 

Now let's look at the Notes from 14% to 16% Yield to Maturity. There are 59 Notes in the same 1,300 or so Notes that have this Yield to Maturity. In this range there is nothing below 1% available. The distribution looks very different than the 16%+ distribution. I stopped at 14% because the number of Notes starts getting larger the smaller Yield to Maturity you get.
If you stack the two graphs you can start to see a boundary of Discount for each Yield to Maturity. I think this is the balance between two Lending Club investment philosophies, Yield to Maturity vs Discount. 19% Y2M Notes seem to sell well at a markup of 2% or less because both types of investors are looking for them (has a low Discount and a high Yield to Maturity). It seems 15% is the low end for markups as high as 2%. I would need more data to be sure (note the three near-zero markup 18.5% Y2M Notes).

This will be one of the things I will start tracking and posting about. When I have more time, I'll delve more into the lower Yield to Maturity Notes.

What are your thoughts?

Update 03/13/2012 @ 12:17pm Central

So I pulled data from the top 1,200 discounted Notes with Yield to Maturity of 6%+. The data is in the graph above. There are a little over 600 Notes from unique Loans that have never been Late, over 6% Yield to Maturity and under 3% markup. The blank spots to the left of the graph probably indicate good markups for quick sales, as there is little competition in that area.

What conclusions do you draw from the chart above?

Saturday, March 10, 2012

Can Someone Check My Math?

(See Other Posts about Lending Club Extension)

In my last post I announced the first release of a Lending Club Chrome Browser plugin I am writing. I am using it myself and making several improvements. One of the improvements is when you offer a Note for sale it will calculate the annualized return range for a given sales price (range because it could be sold any day from today to up to 7 days later). When I was converting my calculation from my Python script, I noticed I was not happy with the original calculation.

I'd really like to have your feedback on the calculation, and have a math check on it.

In precalculus I learned from Ms. Delores Dean that the basic formula when dealing with compounding interest, or rates of return, is:
A = P (R + 1) T
which reads Amount equals Principal times Interest Rate increased by One raised to Time in years. I also wanted to add in Tax, Fees and Payments made. So my final formula was:
S + P (1 - M) - T (S + P (1 - M) - I - FS) - FS = I (G + 1)Y
The variables are as follows:

  • S - Sales Price
  • P - Payments Made
  • T - Tax Rate (ie 0.395)
  • I - Initial Investment
  • F - Fee Rate (ie 0.01) on Sale
  • G - Annualized Gain
  • Y - Years Held
  • M - Maintenance Fee Rate

If you love a simple math problem that requires algebra, please stop reading here and solve for S. If you love a math problem that requires algebra and a little precalculus, please stop reading here and solve for G.

Now an explanation of the equation above. So the actual amount of money you are left with (A in the first equation) is the Sales Price plus Payments Received minus Tax you'll have to pay minus Sales Fee. The Tax would be your Tax Rate (for a conservative estimate of your tax, I recommend using your short term capital gains rate) multiplied by the sum of your Sales Price and Payments Received minus your Initial Investment minus your Fees Paid. And of course the Fee Paid is Fee Rate times Sales Price.

Of note, I assume that the Payments Made needs to be reduced by the Maintenance Fees Paid. Lending Club states that this is 1%, but remember that it is rounded up, usually $0.01 for $25 Notes. I found that the lowest regular payment I've had this month is $0.61, which the fee ended up being 1.64%. So to be conservative, I put the Maintenance Fee (M) at 2% (or 0.02). Some Notes do not have a Maintenance Fee, but to be conservative (and keep things simple) we'll assume there always is one.

Let me know in the comments if there is a problem with this formula or how I am thinking about it.

Now that you've had some time to work through the math, did you get:
G = eloge( (S + P (1 - M) - T (S + P (1 - M) - I - FS)) / I ) / Y - 1
and
S = (I (G + 1)Y - T (I - P (1 - M)) - P (1 - M)) / (1 - T - F + TF)
Let me know in the comments if you got something different.

Now the formula needs to be converted to JavaScript to actually be used in the Chrome Extension. Here is the JavaScript I had for S:
Math.pow(Math.E, Math.log( (salesPrice + (1.0 - maintenanceRate) * payments - taxRate * (salesPrice + (1.0 - maintenanceRate) * payments - investment - feeRate * salesPrice) - feeRate * salesPrice) / investment ) / years) - 1
and for G:
(investment * Math.pow(1.0 + annualGain, years) - taxRate * (investment - (1.0 - maintenanceRate) * payments) - (1.0 - maintenanceRate) * payments) / (1.0 - taxRate - feeRate + taxRate*feeRate)
Anyone see any problems with the JavaScript? Again, please leave a comment.

This is what I am looking at adding to the Sell Notes page using the Chrome Extension. It will add an Annualized Return column and calculate the return if it were sold as soon as possible up to the date selected for the expiration.

This is not a photoshopped image, this is a screen grab from version 0.2 of the extension.

Currently version 0.1 is available, and lets you filter the Browse Notes on FolioFN. When version 0.2 is ready, I'll post again with a list of the enhancements I have made, as well as the assumptions that go into it. I still have a few glitches to work out, but I am excited about 0.2.

Now with 0.2 I can completely ditch my old Python script that used Apple Events to scrape the web pages of Safari to see what I was seeing and calculate these things. Version 0.2 of the Chrome Extension will even one-up the script with a neat little feature that will save me so much time in selling Notes (I'll reveal that in the next post when I release it).

So if you'll do me a favor (and yourself if you plan on using or trying my Chrome Extension version 0.2) and look over my math. I'd really like this to be peer reviewed before I release it.

In the mean time, I'm really having fun using this. I spend so much less time on the FolioFN site and will have fewer human errors.

If I haven't mentioned it yet, please leave me some feedback on the math above.

Thanks.

Monday, March 5, 2012

Alright, I'm Ready To Share

(See Other Posts about Lending Club Extension)

I've been documenting my Lending Club experiece and have been mentioning a script that I use. This script only worked on Mac machines with the Safari browser. I have since started using Google's Chrome browser, and only launched Safari for doing Lending Club business. So to do my Lending Club business in Chrome, I have started writing a Chrome Extension that helps me decide on loans to buy.

This extension is in beta and I make no assurances as to how well it works.

Install Note Browser Filter (click on this link in Google Chrome)

(Note: A slight majority, almost 1/3 of the visitors of this blog are running Chrome, followed by another almost third for FireFox. Also about 2/3 of the visitors are on Windows. I'm hoping this solution works for a majority of the visitors to this site.)

After clicking the install link, there may be a button at the botton of the window asking you to confirm you want to install the extension. Click the "Continue" button and then the "Install" button to install the extension.


This extension will add a button to your toolbar (the red square with the green dollar sign). Clicking this button will reveal extra filters you can apply to the Browse Notes page on FolioFN. You can highlight rows that are out of the acceptable range with whatever color you want. If you specify 000000 as the color, the row will be removed.

For instance, I have it remove duplicate Loans (only the first instance is shown), Loans I already own a Note in, Loans that have Imperfect Status (not Issued or Current), had less than $15 Outstanding Principal or more than $25 Outstanding Principal, Notes that have less the $0.20 of Accrued Interest, Notes that have a Markup of over 10% or Notes that have a Yield to Maturity under 16%.

I also highlight 36 Month Notes in red, Loans that have less than 20 Day Since Payment in orange and Loans with less than 25 Remaining Payments in yellow.

So you sort your Notes the way you always have, and after about 5 seconds it will filter out or highlight the rows for you.

Below is a screen shot of the tool in use. I did not edit this image. Notice it says it is showing Notes 1 - 60, but only 8 Notes are actually showing. Alternating unfiltered rows have green and white backgrounds (to let you know the extension is running, FolioFN uses alternating blue and white backgrounds).
Feel free to leave comments here or email me if you have suggestions or defect reports. I plan on using this on a regular basis, so this is not just something I am throwing out there. I will also be adding features like auto-fill for the Sell Notes page and determining more stats about your performance.