tag:blogger.com,1999:blog-7209606.comments2023-08-24T02:55:54.336-07:00Random ThoughtsAnil Guptahttp://www.blogger.com/profile/04626638497955200142noreply@blogger.comBlogger473125tag:blogger.com,1999:blog-7209606.post-90369177078645685222014-09-15T08:21:20.812-07:002014-09-15T08:21:20.812-07:00Hi,
I came across this article while researching ...Hi,<br /><br />I came across this article while researching the drivers behind LC's Adjustment for Risk and Volatility and the estimates for the assumed default rate for sub-grade A1 and G5.<br /><br />Have you had any followups to this analysis or any insights as to how the differences in estimated defaults between credit grades translate into interest rates?Anonymoushttps://www.blogger.com/profile/06126807193822263171noreply@blogger.comtag:blogger.com,1999:blog-7209606.post-11085161739416003562014-08-17T15:40:50.023-07:002014-08-17T15:40:50.023-07:00anil, thanks for your analysis. As other readers s...anil, thanks for your analysis. As other readers suggest, good stuff. Can we get these graphs in tabular format? want to hook this up with my secondary screens. Anonymoushttps://www.blogger.com/profile/03834514487666776227noreply@blogger.comtag:blogger.com,1999:blog-7209606.post-47019555199996018192014-06-20T13:41:17.028-07:002014-06-20T13:41:17.028-07:00It really worked . Thanks AnilIt really worked . Thanks AnilE336https://www.blogger.com/profile/06230045001764272523noreply@blogger.comtag:blogger.com,1999:blog-7209606.post-77515385037395701562014-05-31T14:49:46.035-07:002014-05-31T14:49:46.035-07:00Here is a formula to consider. Limit a HELOC or RE...Here is a formula to consider. Limit a HELOC or REHELOC to 50% loan to value and it has to be a first lien HELOC. Offer this product knowing that in 10 years the customer may want a Replacement HELOC. Everybody wins, the first investor is paid off after 10 years, the new investor gets a 10 year safe investment that is pretty much economy proof since the LTV is at 50% or less and they are first in line in event of a default.<br /><br />STAY AWAY from second lien HELOC's since those are most apt to defaulting and then there not being enough equity in the home to pay off after the first lien mortgage is paid off. First lien HELOC's mean there is no mortgage other than the HELOC. This also means the homeowner was responsible enough to pay off their home before accessing it for cash in their retirement years.<br />Alessandro Machihttps://www.blogger.com/profile/06316327488702524564noreply@blogger.comtag:blogger.com,1999:blog-7209606.post-1255609811470231992014-05-31T14:44:42.388-07:002014-05-31T14:44:42.388-07:00Peer 2 Peer should have a subset for first lien lo...Peer 2 Peer should have a subset for first lien loans that are backed by collateral. There are millions of homeowners who are equity rich and cash poor and who could be losing their homes even though they have hundreds of thousands of dollars in equity because their HELOC's are resetting and they cannot get a replacement HELOC.Alessandro Machihttps://www.blogger.com/profile/06316327488702524564noreply@blogger.comtag:blogger.com,1999:blog-7209606.post-10138976571141494172014-05-17T23:17:52.947-07:002014-05-17T23:17:52.947-07:00I'am using openstack horizon and I get ambigui...I'am using openstack horizon and I get ambiguity how to insert my android application in this server. I nee some material or any suggestion regarding private cloud using with openstack horizon. my email lakkakulaganesh4@gmail.comAnonymoushttps://www.blogger.com/profile/05788835606775092148noreply@blogger.comtag:blogger.com,1999:blog-7209606.post-23713598522351447052014-05-11T11:42:07.707-07:002014-05-11T11:42:07.707-07:00Anil,thanks for the great posts. Did you attempt t...Anil,thanks for the great posts. Did you attempt to correct for the length of time since issuing? Feels like by the time you ran your analysis many of the loans were only around 12 months and likely didnt have enough time to default yet, making the % that defaults early appear greater than in reality. Cheers, MiguelMiguelhttps://www.blogger.com/profile/07038940716885099106noreply@blogger.comtag:blogger.com,1999:blog-7209606.post-49557299292621826062014-02-04T10:10:16.247-08:002014-02-04T10:10:16.247-08:00Agreed. Private Cloud is the way to go to empower ...Agreed. Private Cloud is the way to go to empower users to spin up/down and manage their own instances and creating a scalable internal infrastructure. And, OpenStack licensing and open source enables it quite well.Anil Guptahttps://www.blogger.com/profile/04626638497955200142noreply@blogger.comtag:blogger.com,1999:blog-7209606.post-38530748217201217042014-02-04T04:54:58.272-08:002014-02-04T04:54:58.272-08:00Each component of the OpenStack has been released ...Each component of the OpenStack has been released under the terms of the Apache license, making it a bit more permissive than the GPL. It also means that you can make any changes under a different license. With an open format and completely modifiable source code, it makes a lot more sense to invest in moving to the Cloud.<br /><br /><a href="http://www.williamsdatamanagement.com/?p=scanning-on-demand" rel="nofollow">Ruby Badcoe @ WilliamsDataManagement</a>Rubyhttps://www.blogger.com/profile/16314434089797506245noreply@blogger.comtag:blogger.com,1999:blog-7209606.post-62173809847937873082013-08-18T19:40:26.128-07:002013-08-18T19:40:26.128-07:00Thanks, so does loan verification matter? Thanks, so does loan verification matter? labohemehttps://www.blogger.com/profile/18116658599393797882noreply@blogger.comtag:blogger.com,1999:blog-7209606.post-74190612090577328022013-07-31T01:20:17.578-07:002013-07-31T01:20:17.578-07:00Hi Anil,
Have you done any regression based analy...Hi Anil,<br /><br />Have you done any regression based analysis to replicate their risk model?<br /><br />thanks,<br />RajeevAnonymoushttps://www.blogger.com/profile/09359593445872476816noreply@blogger.comtag:blogger.com,1999:blog-7209606.post-53955739110661737272013-07-30T04:27:50.046-07:002013-07-30T04:27:50.046-07:00Hi Anil,
Have you done any regressions etc using ...Hi Anil,<br /><br />Have you done any regressions etc using the data available on their web page to duplicate their model of bucketing borrowers into groups A1-D5?<br /><br />thanks,<br />RajeevAnonymoushttps://www.blogger.com/profile/09359593445872476816noreply@blogger.comtag:blogger.com,1999:blog-7209606.post-9441802741723454242013-03-25T00:05:37.116-07:002013-03-25T00:05:37.116-07:00I don't understand why your goal wouldn't ...I don't understand why your goal wouldn't be to maximize ROI. Am i confused, or was Anil basically saying that the opposite of Andrew's strategy is better for making money?K'https://www.blogger.com/profile/15359728563335173146noreply@blogger.comtag:blogger.com,1999:blog-7209606.post-85320999378082491462013-03-24T11:54:53.027-07:002013-03-24T11:54:53.027-07:00Your analysis abilities definitely trump my Excel ...Your analysis abilities definitely trump my Excel pivot table work, but I found that about 80% of loans default within the first 20% of their payments (so 48 months for 60s, and 28/29 for 36s. Thanks for supporting my findings as I use my work extensively (as in automated evaluation of each individual loan) to pick loans which I sort by highest probability of return. Helps to have accurate data when doing that! :-)Kowserhttps://www.blogger.com/profile/04611547744577730175noreply@blogger.comtag:blogger.com,1999:blog-7209606.post-88820232051128562762013-03-22T08:33:16.674-07:002013-03-22T08:33:16.674-07:00"you are taking much higher default risk"..."you are taking much higher default risk"<br /><br />That strategy does indeed incur a higher risk of loss. I agree that it is a fact. However, the degree of increased loss must be weighed against the increased ROI gained from living on the early to mid portions of the amortization schedule over time. <br /><br /><br />"Actually, I think Sam @P2P_NY may have a better strategy of buying loans from secondary market after they have crossed the excess default risk hump in maturity cycle"<br /><br />I suppose it depends if your goal is to maximize ROI or minimize loss. There is indeed a lower risk of loss using that strategy. There is also a lower opportunity of gain. It is great for a conservative investor. I used to do that, but stopped a year ago.Anonymoushttps://www.blogger.com/profile/04995954763070685442noreply@blogger.comtag:blogger.com,1999:blog-7209606.post-36297337852004002862013-03-21T21:15:22.741-07:002013-03-21T21:15:22.741-07:00Thanks for the encouragement Andrew.
I am not so ...Thanks for the encouragement Andrew.<br /><br />I am not so sure about your strategy of turning over notes after extracting the majority of interest. It is not a good strategy considering most of the defaults occur early so you are taking much higher default risk if you are buying new loans and selling them within 12-16 months for 36 month loans. In addition to extra duration risk, and inflation/interest rate risk with 60 month loans, I assume 60 month loans are going to have similar profile of most defaults occurring early.<br /><br />Actually, I think Sam @P2P_NY may have a better strategy of buying loans from secondary market after they have crossed the excess default risk hump in maturity cycle. Basically, buying notes that have negligible or very low risk.<br />Anil Guptahttps://www.blogger.com/profile/04626638497955200142noreply@blogger.comtag:blogger.com,1999:blog-7209606.post-57173056326567402372013-03-21T08:36:22.621-07:002013-03-21T08:36:22.621-07:00Excellent synthesis of data and graphs, Anil! Tha...Excellent synthesis of data and graphs, Anil! Thank you once again!<br /><br />"Can the 60 month loans that continue to make payment make up for this extra loss in principal with longer repayment duration and/or higher interest rate?"<br /><br />Indeed, the search for that answer continues to intrigue me. I lean towards yes, perhaps mostly so for those able/willing to turnover their notes after extracting the majority of the interest.Anonymoushttps://www.blogger.com/profile/04995954763070685442noreply@blogger.comtag:blogger.com,1999:blog-7209606.post-22182865918510526602013-03-19T20:54:24.374-07:002013-03-19T20:54:24.374-07:00Interesting information, thank you for posting!
I ...Interesting information, thank you for posting!<br />I noticed that graph 2 combines 36 and 60 month loans. With 34 months of data for 60 month notes available, is it possible to compare, for instance, the last 30 months of data for 60 month loans with the corresponding past 30 months of 36 month loans to examine the changes in the shapes of those curves over that time frame?<br />That may shed more light on trends between the two maturities.Anonymoushttps://www.blogger.com/profile/04995954763070685442noreply@blogger.comtag:blogger.com,1999:blog-7209606.post-33370244313371168702013-03-16T17:18:59.308-07:002013-03-16T17:18:59.308-07:00Thank you!Thank you!Anonymoushttps://www.blogger.com/profile/04995954763070685442noreply@blogger.comtag:blogger.com,1999:blog-7209606.post-87995177866852431492013-03-16T16:33:41.376-07:002013-03-16T16:33:41.376-07:00Andrew, I updated the second chart in the post to ...Andrew, I updated the second chart in the post to include linear trend line and a screen capture describing the trend model. Thanks. AnilAnil Guptahttps://www.blogger.com/profile/04626638497955200142noreply@blogger.comtag:blogger.com,1999:blog-7209606.post-65424009579591813912013-03-16T15:15:54.440-07:002013-03-16T15:15:54.440-07:00Edited:
For image 2/3, would you mind using the d...Edited:<br /><br />For image 2/3, would you mind using the data to show lines of best fit for the 36 and 60 month scatter plots? Perhaps also an equation that describes each line? If it's not too much work, I'd appreciate it.Anonymoushttps://www.blogger.com/profile/04995954763070685442noreply@blogger.comtag:blogger.com,1999:blog-7209606.post-52379035013018249142013-03-16T15:14:26.024-07:002013-03-16T15:14:26.024-07:00This comment has been removed by the author.Anonymoushttps://www.blogger.com/profile/04995954763070685442noreply@blogger.comtag:blogger.com,1999:blog-7209606.post-84458276134019136222013-03-09T17:28:58.965-08:002013-03-09T17:28:58.965-08:00Lets start with your last statement first. No, th...Lets start with your last statement first. No, the chart doesn't show that earliest 36 month loans was paid off in "actual" month 26.<br /><br />The Months of Payment measurement on the chart is a pseudo-measure instead of actual number of months payment was made. Historical data doesn't provide when loan was paid off so I am using this pseudo-measure months of payment as a proxy. It is important to understand how it is calculated.<br /><br />Months of payments = Total dollar amount paid back / Monthly payment<br /><br />Lets look at an example of a fully paid 60 month loan. I believe the example will also address you second question about there shouldn't be any orange/red star above 36 month mark. I picked a fully paid 60 month loan with Loan ID 638114 for $18,000 loan amount with monthly payment of $440.80.<br /><br />This loan on the chart will show up at intersection of 100% principal paid back and 51.5 months of payment. Actually, this loan was issued 12/31/2010 so it couldn't possibly be paid off after making 51 months of monthly payment. On this loan, borrower made total payments of $22,700.28 that includes all of original principal and interest till the 'actual' month the last payment was made.<br /><br />Months of Payment = $22,700.28 / $440.80 = 51.49 months.<br /><br />Anytime a loan is fully paid off it should show up at 100% principal paid back vertical line and higher than number of months takes to pay off 100% principal horizontal line. In this case,<br /><br />Months of Payment to pay off 100% principal = $18,000 / $440.80 = 40.83 months.<br /><br />Now lets get to your first statement. Loans that are ahead of schedule on payment should show up near top right area while the loans that are behind schedule should show up in bottom left area.<br /><br />Hopefully, this clarifies the confusion, you may have, what I believe between 'actual' months of payment and 'pseudo' months of payment used on these charts.<br />Anil Guptahttps://www.blogger.com/profile/04626638497955200142noreply@blogger.comtag:blogger.com,1999:blog-7209606.post-74480737371224019642013-03-09T08:21:31.005-08:002013-03-09T08:21:31.005-08:00I have some questions about the first chart showin...I have some questions about the first chart showing 36 and 60 month loans.<br /><br />Basically I would expect to see loans that are ahead of the default payment schedule appear below the trend line, and loans that have fallen behind the default payment schedule appear above the trend line.<br /><br />Since the earliest 60 month loans are now hitting the 36 month mark, it seems there should be no red stars above the 36 month line. And the 60 month loans that have paid off early should be indicated by a red bar whose highest data point would appear at the 36 month mark? (The vertical red and blue bars at 100% should show some overlap from the 36 month point down.)<br /><br />Also, to make sure I'm understanding the chart correctly, is it showing that the earliest 36 month note paid off was at month 26? <br /><br />Thanks.<br />flyp52https://www.blogger.com/profile/06850675096574543795noreply@blogger.comtag:blogger.com,1999:blog-7209606.post-27288374969044872412013-03-05T10:30:56.747-08:002013-03-05T10:30:56.747-08:00Grant,
Very unlikely that winter time increase in...Grant,<br /><br />Very unlikely that winter time increase in default rate was due to LC seeking out less creditworthy borrowers late in the year (most probably you are referring to underwriting changes late 2012). Such defaults will earliest materialize in May of 2013. LC will have to be seeking out low quality borrowers in summer of 2012 or earlier for default rate to increase in wintertime.<br /><br />Thanks.<br /><br />AnilAnil Guptahttps://www.blogger.com/profile/04626638497955200142noreply@blogger.com