arby
Member of DD Central
Posts: 910
Likes: 959
|
Post by arby on Nov 19, 2018 14:13:57 GMT
Good point arby, I was trying to work out how the lines will evolve with time and what you've said seems right. That as more loans complete (at whatever stage of their process), the cumulative line will move up at the specific point in the -180 days to infinite days that that happens. So in effect the 2018 line will only start becoming fixed from Jan 1st 2019, at the -180 day point, then -179 day point fixed on Jan 2nd. So would be 180 days into 2019 before we get to the interesting point where the line becomes finalised from the overdue point onwards. Would still be possible to tell if things aren't looking good earlier than that, as the majority of loans will already have been appearing in the data, but nothing final. Thanks. Sums up exactly what I was thinking.
|
|
|
Post by dan1 on Nov 19, 2018 15:21:43 GMT
Good point arby , I was trying to work out how the lines will evolve with time and what you've said seems right. That as more loans complete (at whatever stage of their process), the cumulative line will move up at the specific point in the -180 days to infinite days that that happens. So in effect the 2018 line will only start becoming fixed from Jan 1st 2019, at the -180 day point, then -179 day point fixed on Jan 2nd. So would be 180 days into 2019 before we get to the interesting point where the line becomes finalised from the overdue point onwards. Would still be possible to tell if things aren't looking good earlier than that, as the majority of loans will already have been appearing in the data, but nothing final. I've not got time to reply to all the good points in detail nor reply to the PMs at this moment but just picking up one one point in bold above... In the notes for nerds I had a point which on reflection I should have made more of... - Excludes all active loans with an Expected End date in the future
... which means that the data is valid between -183 day to 0 days overdue even for the 2018 cohort. So, the most worrying part of the 2018 trace is the drop compared to the previous couple of years at the 0 days overdue point.
|
|
arby
Member of DD Central
Posts: 910
Likes: 959
|
Post by arby on Nov 19, 2018 15:28:19 GMT
Good point arby , I was trying to work out how the lines will evolve with time and what you've said seems right. That as more loans complete (at whatever stage of their process), the cumulative line will move up at the specific point in the -180 days to infinite days that that happens. So in effect the 2018 line will only start becoming fixed from Jan 1st 2019, at the -180 day point, then -179 day point fixed on Jan 2nd. So would be 180 days into 2019 before we get to the interesting point where the line becomes finalised from the overdue point onwards. Would still be possible to tell if things aren't looking good earlier than that, as the majority of loans will already have been appearing in the data, but nothing final. I've not got time to reply to all the good points in detail nor reply to the PMs at this moment but just picking up one one point in bold above... In the notes for nerds I had a point which on reflection I should have made more of... - Excludes all active loans with an Expected End date in the future
... which means that the data is valid between -183 day to 0 days overdue even for the 2018 cohort. So, the most worrying part of the 2018 trace is the drop compared to the previous couple of years at the 0 days overdue point. Yes, and and the zero-day point is based on one-third of the amount of data in all the other cohorts. One-third is enough to give an indication of where it's going, that's why it's worrying, but still some time to gain a bit more respectability as the months roll on (or get even worse, could go either way!)
|
|
technik
Member of DD Central
Posts: 72
Likes: 80
|
Post by technik on Nov 19, 2018 16:05:03 GMT
Good point arby , I was trying to work out how the lines will evolve with time and what you've said seems right. That as more loans complete (at whatever stage of their process), the cumulative line will move up at the specific point in the -180 days to infinite days that that happens. So in effect the 2018 line will only start becoming fixed from Jan 1st 2019, at the -180 day point, then -179 day point fixed on Jan 2nd. So would be 180 days into 2019 before we get to the interesting point where the line becomes finalised from the overdue point onwards. Would still be possible to tell if things aren't looking good earlier than that, as the majority of loans will already have been appearing in the data, but nothing final. I've not got time to reply to all the good points in detail nor reply to the PMs at this moment but just picking up one one point in bold above... In the notes for nerds I had a point which on reflection I should have made more of... - Excludes all active loans with an Expected End date in the future
... which means that the data is valid between -183 day to 0 days overdue even for the 2018 cohort. So, the most worrying part of the 2018 trace is the drop compared to the previous couple of years at the 0 days overdue point. That makes sense. But think what I am still saying holds true. Data doesn't include active loans and valid across all timeframes, as thought. But from what I can tell if you update these charts in a month, there will have been loans completing between now and then that fit in the full range -180 to +infinity, some that completed early, some on time, some late, so the line will change wherever they occur. So a loan that initiates on 31 Dec 2018, but then completes a day later (i.e. -180 days) would adjust the -180 point on the line up a bit. My point being the -180 days point can't be forever fixed until 1st Jan. 2nd Jan = -179 day point fixed (because of potential 31 Dec start and 2nd Jan completions). Might have got that wrong though?
|
|
|
Post by dan1 on Nov 19, 2018 16:31:08 GMT
This is great data but one question on 18's data is this based on all 2018 loans or just the ones that should have completed by today if the former then that would explain the difference in the graph if the later then we have a problem brewing? It's based on all loans that should have completed - i.e. the 2018 trace is consistent to the 0 days overdue point but should improve from 0 to 1000 days in time.
|
|
|
Post by dan1 on Nov 19, 2018 18:43:47 GMT
I've not got time to reply to all the good points in detail nor reply to the PMs at this moment but just picking up one one point in bold above... In the notes for nerds I had a point which on reflection I should have made more of... - Excludes all active loans with an Expected End date in the future
... which means that the data is valid between -183 day to 0 days overdue even for the 2018 cohort. So, the most worrying part of the 2018 trace is the drop compared to the previous couple of years at the 0 days overdue point. Yes, and and the zero-day point is based on one-third of the amount of data in all the other cohorts. One-third is enough to give an indication of where it's going, that's why it's worrying, but still some time to gain a bit more respectability as the months roll on (or get even worse, could go either way!) Some numbers for 2018... Loans made live = 628 Loans with an end date on or before today = 317 (i.e. those included in the charts) Loans completed = 179 => 56% of loans completed So, over a half of loans issued this year have an expected end on or before today not the third predicted from the current stage in the year. This is a reflection of the dramatic slow down in loans over the past few months as investors depart. For comparison, the number of loans made live... 2017 = 783 2016 = 605 2015 = 282 2014 = 131 2013 = 28 Note, figures updated today (charts show ~55% generated from data on 10 Nov).
|
|
arby
Member of DD Central
Posts: 910
Likes: 959
|
Post by arby on Nov 19, 2018 19:06:20 GMT
Yes, and and the zero-day point is based on one-third of the amount of data in all the other cohorts. One-third is enough to give an indication of where it's going, that's why it's worrying, but still some time to gain a bit more respectability as the months roll on (or get even worse, could go either way!) Some numbers for 2018... Loans made live = 628 Loans with an end date on or before today = 317 (i.e. those included in the charts) Loans completed = 179 => 56% of loans completed So, over a half of loans issued this year have an expected end on or before today not the third predicted from the current stage in the year. This is a reflection of the dramatic slow down in loans over the past few months as investors depart. For comparison, the number of loans made live... 2017 = 783 2016 = 605 2015 = 282 2014 = 131 2013 = 28 Note, figures updated today (charts show ~55% generated from data on 10 Nov). Half of all 2018 loans to date, but a third of what will likely be all 2018 loans (I specified at the 1 month overdue point). Again, not a criticism, just pointing out that there is still lots more data to come for 2018 (12 times the amount of data for loans that are 90 days overdue)
|
|
|
Post by dan1 on Nov 19, 2018 19:32:31 GMT
Some numbers for 2018... Loans made live = 628 Loans with an end date on or before today = 317 (i.e. those included in the charts) Loans completed = 179 => 56% of loans completed So, over a half of loans issued this year have an expected end on or before today not the third predicted from the current stage in the year. This is a reflection of the dramatic slow down in loans over the past few months as investors depart. For comparison, the number of loans made live... 2017 = 783 2016 = 605 2015 = 282 2014 = 131 2013 = 28 Note, figures updated today (charts show ~55% generated from data on 10 Nov). Half of all 2018 loans to date, but a third of what will likely be all 2018 loans (I specified at the 1 month overdue point). Again, not a criticism, just pointing out that there is still lots more data to come for 2018 (12 times the amount of data for loans that are 90 days overdue) That's not what you posted - you specified the zero-day point... Yes, and and the zero-day point is based on one-third of the amount of data in all the other cohorts. One-third is enough to give an indication of where it's going, that's why it's worrying, but still some time to gain a bit more respectability as the months roll on (or get even worse, could go either way!) FS would have to fill another 325 odd loans before the end of the year if the current number of 317 that should have completed by now is to represent only a third of the total... 317 / (628+325) = 33.3%. The chart is unmistakable in that performance at the zero-day point is the worst of the lot and it doesn't matter which way you dice it.
|
|
arby
Member of DD Central
Posts: 910
Likes: 959
|
Post by arby on Nov 19, 2018 20:06:18 GMT
Half of all 2018 loans to date, but a third of what will likely be all 2018 loans (I specified at the 1 month overdue point). Again, not a criticism, just pointing out that there is still lots more data to come for 2018 (12 times the amount of data for loans that are 90 days overdue) That's not what you posted - you specified the zero-day point... Yes, and and the zero-day point is based on one-third of the amount of data in all the other cohorts. One-third is enough to give an indication of where it's going, that's why it's worrying, but still some time to gain a bit more respectability as the months roll on (or get even worse, could go either way!) FS would have to fill another 325 odd loans before the end of the year if the current number of 317 that should have completed by now is to represent only a third of the total... 317 / (628+325) = 33.3%. The chart is unmistakable in that performance at the zero-day point is the worst of the lot and it doesn't matter which way you dice it. You're right, my mention of the 30-day point was on page 7 of this thread. At the zero-day point, it is still a third of 2018 total if your data was at November 1st. Half of the data will only be reached at Jan 1st. So I approximated and said a third as your data is Nov 10th. None of this was meant to be contentious (the exact opposite in fact), just pointing out the possible impact of early data. i.e. your zero-day completion rate of 35% could actually range from ~15% (if none of the current or yet to be written 2018 loans complete on time) up to ~75% if all remaining 2018 data completes on time. As each day goes on, this range will narrow. I don't disagree that the early signs are very ominous. Edit- for clarity: on 30th June 2018, half of all 2018 loans have been written (I'm assuming equal loans written in each half of the year), but none of those loans are expected to be completed already so there is no 'zero-day' data. On 31st December, all of 2018 loans are written, but only half of them have had expected completion dates, so the zero-day data then has half of the total data that will eventually come to it. Interpolating between, 31st October is a third of what will be the total 2018 data. For completion dates beyond the zero-day mark, there is even less data currently available, but as I've said before, this does not mean the trend will necessarily improve, it could just as likely worsen.
|
|
|
Post by dan1 on Dec 7, 2018 23:40:56 GMT
I've updated the table of Capital Losses, which can be found here. The lost capital to date now stands at £1,193,152. The figure provided on the Loan Statistics page of £1,008,652 was last updated for October and does not yet include the capital losses from the disposal of the Mixed Use Property in Liverpool.
|
|
|
Post by andrews on Dec 8, 2018 15:56:18 GMT
Dear Dan1. I want to just thank you for putting in all that work, to create these statistics of capital losses incurred by FundingSecure. This is most helpful.
|
|
Godanubis
Member of DD Central
Anubis is known as the god of death and is the oldest and most popular of ancient Egyptian deities.
Posts: 2,011
Likes: 1,013
|
Post by Godanubis on Dec 12, 2018 18:23:34 GMT
Not much to attract new investors only 4 new/renewals up for grabs Two are >80% filled.
We need a good few BIG paybacks or things will stagnate very fast.
|
|
coop
Member of DD Central
Posts: 714
Likes: 571
|
Post by coop on Dec 18, 2018 13:17:54 GMT
I've updated the table of Capital Losses, which can be found here. The lost capital to date now stands at £1,193,152. The figure provided on the Loan Statistics page of £1,008,652 was last updated for October and does not yet include the capital losses from the disposal of the Mixed Use Property in Liverpool. Quite incredible that even with their deliberately vague and misleading stats; they STILL cook the books a bit to try and make them look better! Missing out a £185k loss from the latest statistics; one has to believe this didn't happen by accident. fundingsecure you should be ashamed
|
|
Godanubis
Member of DD Central
Anubis is known as the god of death and is the oldest and most popular of ancient Egyptian deities.
Posts: 2,011
Likes: 1,013
|
Post by Godanubis on Dec 18, 2018 13:37:03 GMT
I've updated the table of Capital Losses, which can be found here. The lost capital to date now stands at £1,193,152. The figure provided on the Loan Statistics page of £1,008,652 was last updated for October and does not yet include the capital losses from the disposal of the Mixed Use Property in Liverpool. Quite incredible that even with their deliberately vague and misleading stats; they STILL cook the books a bit to try and make them look better! Missing out a £185k loss from the latest statistics; one has to believe this didn't happen by accident. fundingsecure you should be ashamed What is the loss as a % of total capital and interest returned over the same period ?..
|
|
coop
Member of DD Central
Posts: 714
Likes: 571
|
Post by coop on Dec 18, 2018 13:39:41 GMT
Quite incredible that even with their deliberately vague and misleading stats; they STILL cook the books a bit to try and make them look better! Missing out a £185k loss from the latest statistics; one has to believe this didn't happen by accident. fundingsecure you should be ashamed What is the loss as a % of total capital and interest returned over the same period ?.. I just calculated it and it's... *opens envelope* ... totally irrelevant!!!
|
|