|
Post by chielamangus on Aug 15, 2015 18:29:04 GMT
Dashboard rate 11.1 but XIRR of 8.3 (excluding accruals, as I never count my chickens ...). Cannot understand the latter figure as I was getting 9.5 XIRR last year and I thought I had reduced the investment downtime. No investment under 9.5 nominal.
|
|
|
Post by mrclondon on Aug 15, 2015 18:35:07 GMT
Dashboard rate 11.1 but XIRR of 8.3 (excluding accruals, as I never count my chickens ...). Cannot understand the latter figure as I was getting 9.5 XIRR last year and I thought I had reduced the investment downtime. No investment under 9.5 nominal. By (sensibly) not including accruals, you are by definition not seeing in the XIRR calc any return on capital for any of the distressed loans that you hold that are no longer making monthly interest payments, nor on any interest on manurity loans such as Dorset Dev, Fl** C** etc. (Edit: depending on theproportion of your capital tied up in such loans, it would not be surprising to see XIRR tick downwards on a monthly basis)
|
|
j
Member of DD Central
Penguins are very misunderstood!
Posts: 2,188
Likes: 540
|
Post by j on Aug 15, 2015 23:01:22 GMT
It's gone down since Acne paid out - now only at 14.5% on the dashboard 100% invested since I chomped up some more Aber yesterday, but about 30% is unsurprisingly tied up in suspended loans that may not yield the full coupon. Ah, life on the edge. Don't worry, I too am a grubby rate chaser at 14.6% What % of your portfolio is in distressed loans though chris
|
|
|
Post by chielamangus on Aug 16, 2015 10:34:37 GMT
Dashboard rate 11.1 but XIRR of 8.3 (excluding accruals, as I never count my chickens ...). Cannot understand the latter figure as I was getting 9.5 XIRR last year and I thought I had reduced the investment downtime. No investment under 9.5 nominal. By (sensibly) not including accruals, you are by definition not seeing in the XIRR calc any return on capital for any of the distressed loans that you hold that are no longer making monthly interest payments, nor on any interest on manurity loans such as Dorset Dev, Fl** C** etc. (Edit: depending on theproportion of your capital tied up in such loans, it would not be surprising to see XIRR tick downwards on a monthly basis) Have low percentage of portfolio in suspended loans (8 per cent) only 4 per cent in Dorset, nil in Fl C etc. The 8.3 per cent is on the same basis as the 9.5 from 2014. Actually, just checked the file and the return was even worse - 8.1 !! I guess I now have a tedious job of checking all interest payments. I have noticed that interest in August 2014 was abnormally low, and there was also a substantial dip last October, but the 9.5 of 2014 includes that period. Has there been any event this year which might explain it? Has anyone else such a large difference between the dashboard and XIRR rates?
|
|
Vero
Member of DD Central
Posts: 196
Likes: 163
|
Post by Vero on Aug 16, 2015 15:59:16 GMT
By (sensibly) not including accruals, you are by definition not seeing in the XIRR calc any return on capital for any of the distressed loans that you hold that are no longer making monthly interest payments, nor on any interest on manurity loans such as Dorset Dev, Fl** C** etc. (Edit: depending on theproportion of your capital tied up in such loans, it would not be surprising to see XIRR tick downwards on a monthly basis) Have low percentage of portfolio in suspended loans (8 per cent) only 4 per cent in Dorset, nil in Fl C etc. The 8.3 per cent is on the same basis as the 9.5 from 2014. Actually, just checked the file and the return was even worse - 8.1 !! I guess I now have a tedious job of checking all interest payments. I have noticed that interest in August 2014 was abnormally low, and there was also a substantial dip last October, but the 9.5 of 2014 includes that period. Has there been any event this year which might explain it? Has anyone else such a large difference between the dashboard and XIRR rates?
My dashboard shows 23%. I hold parts with rates from 12% to 18%, but a small amount of loan #35 at 68% is skewing the rate. This rate is for current holdings only, so excludes interest received for redeemed or sold loan parts.
I'm guessing it includes unpaid interest for loan parts that I sold but which were later suspended, as they still show in my "loan book" (I retain a £1 holding of sold parts, to keep track of unpaid interest).
I have withdrawn over 90% of this account this year so this rate is only representative of the leftovers, so not very useful to me.
As the rate relates mostly to unpaid interest which I do not really expect to receive, I have not paid it much attention and have not used XIRR.
|
|
|
Post by chris on Aug 16, 2015 16:39:26 GMT
Don't worry, I too am a grubby rate chaser at 14.6% What % of your portfolio is in distressed loans though chrisHaven't calculated it but it'll be higher than average.
|
|
|
Post by chielamangus on Aug 16, 2015 17:02:41 GMT
The mystery of my low XIRR has been solved! Perhaps it's a cautionary tale. I broke down the sequence of payments into smaller discrete periods of one month, and tested each month to see where any aberrations might occur. it soon became obvious that a substantial withdrawal in October followed by some deposits created a situation that the XIRR algorithm could not handle. I therefore used XIRR to calculate returns for subsequent periods and even then for November and December I got some very strange results indeed - 0.23 per cent for November and 0.34 per cent for November & December combined. The simple returns in these periods were far greater than this. I then estimated for 2015 to end July and the XIRR was 11.18 per cent - remarkably similar to the weighted average return given on the dashboard.
Searching the internet I discover that XIRR has problems computing certain types of cash flow, & I'm not the first to find it can produce nonsense results. I found a site (Pine-grove.com) which allows one to compute an IRR for up to 24 items of income or expenditure, so checked the XIRR results against theirs. The November figure was 17.84 per cent and the November-December figure was 14.56 per cent. The 2015 figure was exactly the same as XIRR at 11.18 per cent.
My version of Excel is 2003 so perhaps MS have done something about the algorithm since then. But all this is a warning not to put blind faith in XIRR.
[NB In calculating the IRR I use only the initial investment (or AC balance at the beginning of the month), transfers to and from my bank account, and the AC balance at the end of the month.]
|
|
jonno
Member of DD Central
nil satis nisi optimum
Posts: 2,808
Likes: 3,242
|
Post by jonno on Aug 17, 2015 9:16:25 GMT
Dashboard rate 11.1 but XIRR of 8.3 (excluding accruals, as I never count my chickens ...). Cannot understand the latter figure as I was getting 9.5 XIRR last year and I thought I had reduced the investment downtime. No investment under 9.5 nominal. By (sensibly) not including accruals, you are by definition not seeing in the XIRR calc any return on capital for any of the distressed loans that you hold that are no longer making monthly interest payments, nor on any interest on manurity loans such as Dorset Dev, Fl** C** etc. (Edit: depending on theproportion of your capital tied up in such loans, it would not be surprising to see XIRR tick downwards on a monthly basis) mrclondon; describing a proportion of AC's loanbook as MANURITY loans, has to be one of the greatest Freudian Slips I have ever encountered! Brilliant!
|
|
upland
Member of DD Central
Posts: 479
Likes: 175
|
Post by upland on Aug 18, 2015 7:55:31 GMT
I think that it is probably a difficult thing to work out meaningfully , FC quotes three figures that most of us would be used to. I have a newish holding of about 60 loans , I started 3-4 months ago with a bit over 2 K now. The dashboard rate varies from 9.9 to 10.15. I have 15% combined in the GBBA and GEIA (both equal). Typically I have weighted amounts with much more in the loans that I like the look of. But as I am no expert so a little in most of them hedges my bets a bit I feel. It makes admin and searching a bit easier for me as the risked sum indicates what I thought about it at the time. I have about 15% cash but as I grow the holdings intend to keep that to less than 5%. Most of the holdings have come from shrapnel.
I put more money into simple loans that I understand with property type first charges , avoid very high LTVs , most environmental loans or ones with only a month or two to go. I put less in second charges , trade debenture types , low interest return propositions , multiple interconnected situations , or these loans that have an LTV based upon what it will be worth when finished.
It seems to have taken a long time for much income to trickle onto the overall sum.
I dont expect it to settle down for some months more but eventually I should be able to estimate monthly income and work out some sort of real return. As I see it even with asset security when a loan goes wrong one will probably be deprived of the use of the capital for perhaps up to two years and so the headline return rate is impacted significantly by relatively few bad loans. So I am hoping that the fluctuation in the return will be fairly small.
|
|
|
Post by chris on Aug 18, 2015 12:07:21 GMT
As I see it even with asset security when a loan goes wrong one will probably be deprived of the use of the capital for perhaps up to two years and so the headline return rate is impacted significantly by relatively few bad loans. So I am hoping that the fluctuation in the return will be fairly small. Whilst this is true you would also be "earning" default interest which may or may not be recoverable as well, depending on how good the security is. We've got plans to introduce XIRR figures to the dashboard, both with and without accrued interest being included, which will help this type of analysis. There are a couple of higher priority jobs to come first so it's likely to be another month or so at the earliest.
|
|
upland
Member of DD Central
Posts: 479
Likes: 175
|
Post by upland on Aug 18, 2015 13:12:37 GMT
As I see it even with asset security when a loan goes wrong one will probably be deprived of the use of the capital for perhaps up to two years and so the headline return rate is impacted significantly by relatively few bad loans. So I am hoping that the fluctuation in the return will be fairly small. Whilst this is true you would also be "earning" default interest which may or may not be recoverable as well, depending on how good the security is. We've got plans to introduce XIRR figures to the dashboard, both with and without accrued interest being included, which will help this type of analysis. There are a couple of higher priority jobs to come first so it's likely to be another month or so at the earliest. Thanks Chris , I was hoping that with AC the quality of the disaster when it goes wrong would be better than from elsewhere. It seems easy to lose 25-30% of your return with loans that go wrong hence my diversified and tortoise like approach. Surely with the higher yields the number of loans one holds must be less and any interruption will dent the returns more significantly ?
|
|
duck
Member of DD Central
Posts: 2,879
Likes: 6,956
|
Post by duck on Aug 19, 2015 4:38:35 GMT
... My version of Excel is 2003 so perhaps MS have done something about the algorithm since then. But all this is a warning not to put blind faith in XIRR. ..... I use later versions (mainly 2007) and this 'fault' is still present. As an investor in Bondora (drawing down at present) where return is heavily influenced by a heavy default rate and poor recovery I started incorporating my own 'actual returns' calc to my spreadsheet - something that I now use for for all platforms. All incoming payments are taken into account (returns % rate changes by the day) but a figure is frozen for each month end to help identify trends.
Currently XIRR shows returns 5.1% higher than my 'actual return' with Bondora.
With my 'actual return' calc I also take into account Tax due to the current rules of tax on all income* and no offsetting of losses, something that XIRR also struggles with. I use a different calc for my business accounts and a slightly different version with my estimated losses included .... that way I have a best and worst figures on a daily basis.
*Personal not business accounts. Not seen any changes being brought forwards since I posted p2pindependentforum.com/thread/2447/peer-relief-proposed-technical-criteria anybody heard anything?
|
|