-
12-09-2018, 08:37 PM
#3671
yeah, nah
Age Spread of Loans
The chart below shows the age spread of both my current loans and loans that have defaulted.
Orange: Age spread of current loans - scale on the left axis (example: there are 130 loans that are 15 months old)
Blue: Age spread of defaulted loans - scale on the right axis (example: 4 loans defaulted 6 months after purchase)
Age of Loans.png
I have no clue why month 8 and 9 are so low in relation to loans defaulting - just an anomaly due to low number of overall defaults at this point in time I guess.
An informative graph I think - for my loan set it shows that a good portion (~35%) are now out of the 'default' danger zone.
Notes:
- Loan default date is taken as charged-off date, not last payment date.
- Three of the charged-off loans are currently still being paid...Harmoney data errors......
- This would be a great graph to have time-lapsed, but it's just too much work to be bothered...
-
13-09-2018, 01:51 AM
#3672
yeah, nah
Current / Paid Off Loans
Another graph that helps paint a picture of loan characteristics:
Orange: Age spread of current loans.
Green: Age of paid off loans.
Current - Paid Off Loans.png
So a large portion of my loans are repaid at the 4th month through to at least the 8th month.
Notes:
- Around 80% of my loans are 60 month term loans
- In 18 months I have purchased over $220,000 worth of loans from a total deposit of $100,000 (no withdrawals)
- The last interest rate adjustment may have had a significant impact on the timing of paid off loans?
- It's worth noting that the original $100,000 was invested in the first 3 to 4 months, all loans purchased after that have been paid from $'s returned from paid off loans and interest.
- I have been increasing loan size in recent times (~ 4-6 months) to help keep up with re-investment (not represented in the graph - I should do the same graph but use $ value?)
- Around 35% of my original $100,000 loans have progressed past 14 months!
Last edited by myles; 13-09-2018 at 08:35 AM.
Reason: additional notes+fix chart
-
13-09-2018, 10:37 AM
#3673
yeah, nah
Something that I found when looking through my data set is that none of the re-written loans have defaulted!
This could be partly due to timing and the limited time frame of my loan set, but I find it very interesting.
Cool Bear, is this something you could chart with your extended data set? Just the same as your previous default graph, but only include loans that have previously been re-written i.e. positive value for Previous Loan Pay-off (re-write) column.
-
13-09-2018, 10:46 AM
#3674
Originally Posted by myles
Something that I found when looking through my data set is that none of the re-written loans have defaulted!
This could be partly due to timing and the limited time frame of my loan set, but I find it very interesting.
Cool Bear, is this something you could chart with your extended data set? Just the same as your previous default graph, but only include loans that have previously been re-written i.e. positive value for Previous Loan Pay-off (re-write) column.
30 Months ago we opened a HM account for my wife.
Specifically we invested Only in Re-Writes and also only in the Maximum Re-Write allowed for that Grade of loan.
Pleased to report that No Loan has every Defaulted or been Over due with payment over that time.
-
13-09-2018, 11:49 AM
#3675
Member
Originally Posted by Saamee
30 Months ago we opened a HM account for my wife.
Specifically we invested Only in Re-Writes and also only in the Maximum Re-Write allowed for that Grade of loan.
Pleased to report that No Loan has every Defaulted or been Over due with payment over that time.
I dont have any Write offs on rewrites - I do however have a number of arrears , hardship, protect waivers etc
Including some that are 180+ days in arrears
Attachment 9924
-
13-09-2018, 08:57 PM
#3676
Member
Originally Posted by myles
I take the comments above, but the point I'm trying to show is the rate/value of defaults per grade. Defaults are not annual, they are total over the life of all loans - the %Loss is based on total invested value per grade, not current or final value (so loss of potential interest is not included).
I take Cool Bears point on fees, so I've added that in - it had no effect on the overall trend, but it could have. Tax is at a portfolio level so I'm not including it deliberately.
I know the last column is meaningless, but I find it to be indicative of the return for the grade.
Updated with 15% loss due to fees:
Attachment 9918
The key thing I take from these values is that the expected, larger default losses for higher grades is not what I'm seeing. So selection criteria can impact expected defaults and averages - significantly.
Further to Coolbear and Snow Leopard, That analysis although interesting overlooks much. Defaults, will create a drag on your return - due to the capital not being recuperated so interest is foregone.
Fee should be applied against all interest.
Early repayment amplify's the impact of default, especially in the higher grades. Without the benefit of your data, would say your E5 grade is returning closer to 12% rather then 19%. See attached.
Capture.JPG
That said your DEF grades are performing very nicely, and well below the predicted static loss ranges indicated by HM, so you must a knack for risk selection
-
13-09-2018, 10:56 PM
#3677
yeah, nah
Originally Posted by leesal
Early repayment amplify's the impact of default, especially in the higher grades. Without the benefit of your data, would say your E5 grade is returning closer to 12% rather then 19%. See attached.
I re-invest 'early repayments' - so shouldn't the months be 14 with the associated increased interest and fees?
This wasn't the point I was trying to highlight. The point was the %Loss column - for my loans it is clearly more advantages to invest in higher grade loans based on loss due to defaults (when interest is taken into account). Compare the %Loss column for B5 vs D4 or B4 vs E3 and then consider what the return for each will be... Higher grade loans, for my loan set, are not showing increased default losses as would typically be expected (include the interest gain and it should be obvious how significant the gain is).
-
14-09-2018, 07:23 AM
#3678
Member
Originally Posted by myles
I re-invest 'early repayments' - so shouldn't the months be 14 with the associated increased interest and fees?
The reinvestment principal is independant, so in the case of E5 - if you were to reinvest the returned capital back into a E5 loan it should have the same characteristics and return 12% in this example (eg expected default of 4.5% per year, 35%pa of early repayment, etc).
Originally Posted by myles
This wasn't the point I was trying to highlight. The point was the %Loss column - for my loans it is clearly more advantages to invest in higher grade loans based on loss due to defaults (when interest is taken into account). Compare the %Loss column for B5 vs D4 or B4 vs E3 and then consider what the return for each will be... Higher grade loans, for my loan set, are not showing increased default losses as would typically be expected (include the interest gain and it should be obvious how significant the gain is).
Agree with that, you are showing incredible risk selection at the DEF grade. I find it incredible that your D grades are outperforming your C's. Your E grades are showing a static loss of 4%, HM are showing a static loss of 7.7% on the 2017 cohort. So you are doing 50% better then the platform, a impressive record!
-
14-09-2018, 09:39 AM
#3679
-
14-09-2018, 10:51 AM
#3680
Member
Originally Posted by myles
Something that I found when looking through my data set is that none of the re-written loans have defaulted!
This could be partly due to timing and the limited time frame of my loan set, but I find it very interesting.
Cool Bear, is this something you could chart with your extended data set? Just the same as your previous default graph, but only include loans that have previously been re-written i.e. positive value for Previous Loan Pay-off (re-write) column.
Will try but have to be in a month or two as I will be travelling.
Tags for this Thread
Posting Permissions
- You may not post new threads
- You may not post replies
- You may not post attachments
- You may not edit your posts
-
Forum Rules
|
|
Bookmarks