Neil_P2PBlog
P2P Blogger
Use @p2pblog to tag me :-)
Posts: 355 
Likes: 209
|
Post by Neil_P2PBlog on Aug 19, 2016 13:56:34 GMT
I'm trying to build a tax calculator for buying late loans on the secondary market. Its straight forward to put together the factors for: - Accrued tax liability at e.g. 20% tax band based on days active/left
- Capital gain from buying at discount
- Annualised % from discount%/ (days remaining / 365)
What I am struggling with is accounting for the increased risk of default/ bad debt. From the statistics page I can come up with 2 different ideas of what bad debt % might be: - Looking at last 3 months, subtract around 12.48% expected return from around 12.26% actual return to give bad debt of around 0.22%
- Looking at their predicted rates, subtract 12.7% gross interest payable from 11.2% estimated net return to give bad debt of around 1.5%
Once I have a number for this, I'll multiply it by (length of loan term/ length remaining). My thinking for this is if you are buying 180day loans with 90days remaining you would have twice the bad debt risk, and 180day loans with 45 days left you'd have four times the bad debt risk. Obviously loans with higher interest rates have higher default risks but I can't find numbers to build that in. Once I have the numbers above I can just sum: A) £x gain from buying at discount B) £x loss from taking on tax liability C) £x loss from expected increased risk |
|
SteveT
Member of DD Central
Posts: 6,873
Likes: 7,918
|
Post by SteveT on Aug 19, 2016 14:22:04 GMT
Not sure that your logic makes sense. A loan on FS is only likely to reveal repayment problems at / near the end of the 6 months term. If you intend to hold a loan to term then there's little difference in default risk between buying it with 183 days to run or 31 days.
However, a loan you hold to term has an infinitely higher risk of giving you default issues than one you sell sometime before term.
|
|
Neil_P2PBlog
P2P Blogger
Use @p2pblog to tag me :-)
Posts: 355
Likes: 209
|
Post by Neil_P2PBlog on Aug 19, 2016 14:50:04 GMT
Not sure that your logic makes sense. A loan on FS is only likely to reveal repayment problems at / near the end of the 6 months term. If you intend to hold a loan to term then there's little difference in default risk between buying it with 183 days to run or 31 days. However, a loan you hold to term has an infinitely higher risk of giving you default issues than one you sell sometime before term. I agree that a loan is likely to default only at the end/near end, and that there is no risk of default if you manage to sell early. Let's just say a figure of 1.5% for losses due to bad debt overall on FS. Strategy 1: Maintain an investment of 50x £100 loans. You buy all loans new and hold to term. During a year you have 100 different loans and would expect that 1.5% of losses we assumed above. Strategy 2: Buy discounted loans on the secondary market half way through their term, with the same total investment as strategy 1. Now you have 200 different loans during the year with the same chances of bad debt over the entire loan term. My logic is that holding twice as many loans through the risky period your expectation of bad debt is now 3% in relation to your original investment. Strategy 3: Buy discounted loans with 1/4 of term to go. You now hold 400 different loans during the same year and are likely to experience 4x the number of defaults in 1, so 6%.
|
|
SteveT
Member of DD Central
Posts: 6,873
Likes: 7,918
|
Post by SteveT on Aug 19, 2016 14:59:07 GMT
Nothing is repaid on FS until a loan completes, so why would you have more loans under your 2nd and 3rd strategies (assuming you buy the same amount in each loan)? In fact you'd need to spend more to buy the same £ amount since you also need to pay the seller their accrued interest, less any discount.
|
|
Neil_P2PBlog
P2P Blogger
Use @p2pblog to tag me :-)
Posts: 355
Likes: 209
|
Post by Neil_P2PBlog on Aug 19, 2016 15:18:23 GMT
The % bad debt loss and interest %'s are all annualised numbers, so you need to recycle your investment income on new loans to maintain your £invested. As in strategy 2, if you are buying 6 month loans after 3 months, you'll have to buy twice as many loans over the year to maintain your investment and gross % return.
|
|
duck
Member of DD Central
Posts: 2,578
Likes: 5,698
|
Post by duck on Aug 19, 2016 17:35:31 GMT
... Strategy 3: Buy discounted loans with 1/4 of term to go. You now hold 400 different loans during the same year and are likely to experience 4x the number of defaults in 1, so 6%. Whilst this approach appears logical take it a few more iterations down the line and at 1/64th of term (OK you can't buy / sell at that stage but bear with me!) you are up to 96% default rate so take it one step further and you are well over 100% which is mathematically 'strange'. With suitable diversification I have always worked on a pessimistic 1.5% of cash loss against cash invested I apply this to all platforms that I invest in not just FS. So far I am yet to even vaguely approach that level on any platform in any year. Yes there have been some sweaty palm moments and some lucky escapes but that is all part of the process. I note Neil_P2PBlog that you haven't mentioned 'recoveries' (I've had several with FS recently) these skew overall returns as well .......
|
|
Neil_P2PBlog
P2P Blogger
Use @p2pblog to tag me :-)
Posts: 355
Likes: 209
|
Post by Neil_P2PBlog on Aug 19, 2016 17:58:59 GMT
... Strategy 3: Buy discounted loans with 1/4 of term to go. You now hold 400 different loans during the same year and are likely to experience 4x the number of defaults in 1, so 6%. Whilst this approach appears logical take it a few more iterations down the line and at 1/64th of term (OK you can't buy / sell at that stage but bear with me!) you are up to 96% default rate so take it one step further and you are well over 100% which is mathematically 'strange'. With suitable diversification I have always worked on a pessimistic 1.5% of cash loss against cash invested I apply this to all platforms that I invest in not just FS. So far I am yet to even vaguely approach that level on any platform in any year. Yes there have been some sweaty palm moments and some lucky escapes but that is all part of the process. I note Neil_P2PBlog that you haven't mentioned 'recoveries' (I've had several with FS recently) these skew overall returns as well ....... I just call it 'bad debt', but ideally I'm looking for a % that includes recoveries. The cash loss becomes confusing with later term loans, as the accrued interest becomes your cash loss, so is not included in the .07% historic capital loss FS quote. Great observation on taking it a few more iterations, I need to do more thinking on that one!
|
|
duck
Member of DD Central
Posts: 2,578
Likes: 5,698
|
Post by duck on Aug 20, 2016 5:33:02 GMT
Whilst my approach doesn't give you an answer to your question I can outline how I work out my current position, it might give you a few pointers
I have a running weighted interest % rate (interest rate for each loan and amount invested in each loan)
I have a running amount invested (this varies as loan numbers vary up and down)
I have a running amount un invested
Then I then work out 5 figures
Projected interest (no losses)
Projected interest less 1.5% cash invested
Interest received
Total amount invested (takes into account cash inputs/withdrawals)
Value of account less defaulted loan amounts
When a loan is in default (or has any issues that give me serious doubts that I will see my money again) it is removed from all 3 running figures ..... treated as 100% loss.
When recoveries are made these are fed back into interest received and value of account. (this is an automatic action inside the spreadsheet) This is a pessimistic approach but for me it gives meaningful graphs that I can see trends which is very useful especially when money is moved from platform to platform ....... I can always see the actual deviation between ideal interest and my 'notional' 1.5% capital loss by using the account vaules.
At the end of each calendar month the figures are 'frozen' which allows a graph to be produced to show/preserve trends.
This sounds like a lot of work but the spreadsheets for each platform read from the downloaded data from each platform so my only actions are to keep the loans 'tab' current with which loans I am actually invest in and their interest rate.
The one exception to this approach is Bondora (notoriously high default rate) where I monitor defaults specifically (currently have 899 loans in default!) and I monitor defaults and recoveries on a stand alone basis again freezing the position at each month end.
Hope you can glean some pointers from the above ......
|
|
