Harmoney

[myles]

I'd be interested to hear what you are doing in this space?

I've covered what I do/have done a while back - the most basic was to write of any portion in arrears over 31 days - at the time I prefered to over estimate losses (still do). I still track that so I know if a big hit is likely to come through - but I'm not seeing this is really necessary, but nice to have. With over 100K, defaults, at the current rate I'm getting them, are looking stable. I'm expecting some growth in defaults but predict compounding will keep up with it at the very least. If that changes I'll see it coming.

I'm in a unique situation that I've invested 100K quite quickly (3-4 months) and have not withdrawn, so I'm getting quite a good picture of what's going on without being tainted by deposits/withdrawals. I'm surprised at, what is currently, a rock steady straight line of growth. Perhaps I've just hit on a reasonable loan selection process and range? [I gave up trying to get an accurate figure when the new rates etc. came in - not worth the effort for me - started up another account to be a little more conservative, but have pretty much fallen back to the same as my original account as it's doing well and looks like the sweet spot I'm after and still fit's with the new rates].

One thing I did in the lead up to Christmas was to tighten up my selection process as it's been shown that P2P is similar tos Credit Card default rates - this period being a higher default rate period. This has created a bit of a 'wobble' as I couldn't match enough loans and my available principle crept up to >10K (it comes back quickly!), working on getting most of it back in now. It is a finicky beast, with so many influences...

Moral Hazard - its an insurance term. The more dubious the policyholders credentials, the higher the cost from moral hazard.

Aware of what it means - not sure how PP at a lower rate, is different from PP at a higher rate, hence my query.

Its quite a basic investment concept. Just google. If you've are paying into kiwisaver you are already dollar-cost-averaging.

Aware of the term, doesn't seem to match what you are doing i.e. investing in whatever comes along vs what you've indicated - that you are quite selective...

[leesal]

I've covered what I do/have done a while back - the most basic was to write of any portion in arrears over 31 days - at the time I prefered to over estimate losses (still do). I still track that so I know if a big hit is likely to come through - but I'm not seeing this is really necessary, but nice to have. With over 100K, defaults, at the current rate I'm getting them, are looking stable. I'm expecting some growth in defaults but predict compounding will keep up with it at the very least. If that changes I'll see it coming.

I'm in a unique situation that I've invested 100K quite quickly (3-4 months) and have not withdrawn, so I'm getting quite a good picture of what's going on without being tainted by deposits/withdrawals. I'm surprised at, what is currently, a rock steady straight line of growth. Perhaps I've just hit on a reasonable loan selection process and range? [I gave up trying to get an accurate figure when the new rates etc. came in - not worth the effort for me - started up another account to be a little more conservative, but have pretty much fallen back to the same as my original account as it's doing well and looks like the sweet spot I'm after and still fit's with the new rates].

One thing I did in the lead up to Christmas was to tighten up my selection process as it's been shown that P2P is similar tos Credit Card default rates - this period being a higher default rate period. This has created a bit of a 'wobble' as I couldn't match enough loans and my available principle crept up to >10K (it comes back quickly!), working on getting most of it back in now. It is a finicky beast, with so many influences...

An elegant way of treating, that will give you a conservative view; What is the payment rate of 31-60 that you are finding so far? Are you taking the data from the "export" button in reports, or doing your own ageing?

Aware of what it means - not sure how PP at a lower rate, is different from PP at a higher rate, hence my query.

Some insurers use credit risk as part of their underwriting criteria. This is due to the relatively high correlation between loan default and claims incidence (through moral hazard - eg using imdemnification to ones advantage to engage in risky behaviour).

So - imagine holding a $20000 Harmoney loan with PP, in the knowledge that layoffs are coming...

The thought process of a borrower who has diligently settled all their past obligations (grade A), vs somebody with a track record of defaulting (grade E/F) may lead to different behaviours. Thus E/F on average showing an increased incident in envoking payment protect. Would expect this to hold true here.

[Cool Bear]

I am sitting abut 1% lower than you after about 3 years/6000 loans . I am interested in gaining some understanding of the difference ie due to higher risk adopted or better selection of loans, or a combination. Do you read the stories or just rely on filters? I essentially rely on autolend and filter out all business loans, all loans above $35k and all A's, E's and F's and 25% income ratio. Until 1.5 I also invested in A's and had a roughly equal weighting in each major grade. Any insights would be most welcome.

RMJH, it is very difficult to see what works and what does not. Last year, I started analysing my 2016 loans as a group. The idea is so that new loans do not affect the analysis. Also my investment criteria change as time goes by but did not for that year. Harmoney's interest rates also did not changed for those 12 months as well. The objective was to see what grades gave me the best returns. I started tracking that result as each month goes by in 2017. That analysis is now worthless with Harmoney version 1.5 as a borrower who qualify under say C3 previously may not be assigned the same grade under V1.5.

Most of my loans now are via Autolend (B to E) so I do not get to read the stories before the loans are committed. With the manually done loans, I do glance at their stories (borrower's comments). However, income ratio and grades are still more important. I do loan to all grades A to F manually. With manual loans, marital status also affects my decision a bit as "divorce" and "separated" complicates life for the borrowers. I do not spend much time on deciding, most manual loans will take me about 2 to 3 seconds and at the very most 10 (very rarely).

[alistar_mid]

just did some number crunching on profitability by loan type for my portfolio.

its all in the attached excel file, everything’s hard coded except the most recent date which is linked to a csv download, so people can critique the formulas / logic. (yep I could do it with pivot tables but i like sumifs)

so profit and loss for me is pretty straight forward, I claim write offs and fees, so for me its gross interest less fess less writes offs = gross profit, less tax = net profit.
So for each grade I have a net profit.

The profitability is harder to work out. Obviously it’s the profit / investment for each grade.

However I have chucked money in at different times and chosen loans at different times. So how I have done this is worked out an average length of loan, which you can get by taking for all the loans in status arrears or current, by taking their loan term less payments remaining. This should give the term the loan has been “active”. You get a total months figure for each loan grade, and then you divide by loan counts to get the average length of loan.

Then to work out average investment, I have summed amount invested, and scaled it by average loan length, to annualise it. For example I have dropped $11,300 into A’s over the life of my portfolio, the average length of an active A is 8.62 months, so to annualise or weighted average the amount I have invested in A’s to get an annualised return, its $11,300 * (8.62/12) = $8,133

A’s have made me a net profit of $464 (this is from the life of my portfolio) so the roi on A’s is $464 / $8,133 = 5.70%.

I’m not sure of the logic. The profit and loss is reflective of the life of my portfolio, which is 16 months. When working out the average amount invested, I’m trying to annualise it.

Not sure it matters too much, the results will be the same, D’s are best for me, followed by C’s.

.Attachment 9431


hmm it appears I can;t share my spreadsheet

[beacon]

so people can critique the formulas / logic.... hmm it appears I can;t share my spreadsheet

Interesting concept Alistar. Segregating and analysing v1.5 only loans (rather than your to-date totals) might provide you purer insights by grade and be more useful.

[alistar_mid]

Interesting concept Alistar. Segregating and analysing v1.5 only loans (rather than your to-date totals) might provide you purer insights by grade and be more useful.

oh yeah right, apples and oranges.

good point.

[beacon]

for all the loans in status arrears or current, by taking their loan term less payments remaining. This should give the term the loan has been “active”. You get a total months figure for each loan grade, and then you divide by loan counts to get the average length of loan...

Excluding Paid-offs from calculation can only give you expected return from your current portfolio - rather than actual overall. Also, use of months and Harmoney classification thereof should increase margin of error.

[mccollr]

AI assists in assessing credit risk.

https://www.cio.co.nz/article/632369...s-credit-risk/

[BJ1]

Just as a matter of interest people: when looking at your export sheet, what proportion of live loans which should have received a payment recently are recorded as Current but actually have had no payment this past month - say up to and including the 15th of the month. I have 4% of my live loans recorded as Current which are actually in arrears. That's on top of the 2.5% recorded as in arrears. Anyone else chatting to Harmony about this?

[777]

Just as a matter of interest people: when looking at your export sheet, what proportion of live loans which should have received a payment recently are recorded as Current but actually have had no payment this past month - say up to and including the 15th of the month. I have 4% of my live loans recorded as Current which are actually in arrears. That's on top of the 2.5% recorded as in arrears. Anyone else chatting to Harmony about this?

Probably reflects the delay in actual payment and you actually being credited with it. Usually 2-3 days.