|
Post by markimus on Jul 28, 2020 15:54:44 GMT
Yep it makes sense, when they scrapped the markets for new lenders it was always a question, pretty much an assumption, that at some point they would turn off all new lending to 1 and 5 as well. Just like they did with 3. I assume Covid just brought this forwards, and with new markets generally being at lower rates it seems silly they would continue to lend in the old ones. I suspect everyone will lose access to 1 and 5 at some point. You can probably delay for now but I think they will naturally die as the last loans come to an end. Yes agreed it makes sense given the phasing out, but at least one user on here has recently reported a new loan in their 1 Year account with the full 12 months remaining, which would suggest a new loan in the market (I'm afraid I can't remember exactly who said it and where). Perhaps this is occurring still in 1 Year, but not 5 Year (hence faster movement in 5 Year?). It's more just that the splitting up of the markets for RYI purposes makes somewhat less sense to me in light of this. Anyway, nothing that can be done about it. (Although I'd still like to see that RYI release breakdown!) I wonder if the lending is turned off because they are only trying to service existing agreements (for example giffgaff, which I'm assuming are small loans and are fullfilled by A/P/M markets) and the 5 year market was always for big personal loans? Another option could be they're trying to clear the 5 year market completely and to try and get some lending resummed in the future?
All just assumptions of course, I'm just curious what their plans are if the metro deal falls through.
|
|
aju
Member of DD Central
Posts: 3,496
Likes: 923
|
Post by aju on Jul 28, 2020 16:09:13 GMT
At least they acknowledged it was business decision, sadly I'm not sure it will help except I would say that 5Y and 1Y are getting closer for us every day, at least they are where our most money is. We still have one account that has quite a bit of access though. We are getting quite a bit with loans closing too. I'm guessing many were due but have not checked as not really that bothered as long as I don't miss them and inadvertently end up lending more by mistake. As I said elsewhere the new lend on all accounts seems to be restricted to Max 8% but the relend so far is still 10% but for how much longer that will last is not clear. Sorry, aju, I'm not sure I follow the relevance of this to what I was posting about... I was initially responding to the comments RS made about the decision not to present any info on ryi split across products. I was explaining why I don't feel the need to chase RS - I left out the critical reason in that I don't think its worth the bother as they probably won't decide to share the info. (It wouldn't be that hard for them to share this should so wish to share in the future though!. My other comments were not relevant but I was sharing that too, just because I can ...
|
|
|
Post by RateSetter on Jul 28, 2020 16:09:28 GMT
Good afternoon. Today we have delivered £0.6m, and the full update is below:
|
|
aju
Member of DD Central
Posts: 3,496
Likes: 923
|
Post by aju on Jul 28, 2020 16:22:30 GMT
I've noticed that Trust Pilot for RS is posting a "great" status again but still RS site says Excellent!. I've reported it today but after reporting it in middle of July it seemed to raise itself back up with a few good reviews but over the last 2 days or so the reviews are getting very testy again. Perhaps some of the people here are reporting over there although it's not easy to tell. RS keeps saying that the Graphic where they declare the TP results on their site is automatically copied across but I'm not sure that is the case. Back in middle of July it was only when I flagged it did it suddenly change back but then they claimed it was because the TP scoring rose up again - there were some odd, almost suspect 5* 1st timer votes that I guess that pulled it back before they addressed the issue.
|
|
beagle
Investor in ratesetter, funding circle, lendy (lesson learnt) and AC
Posts: 670
Likes: 322
|
Post by beagle on Jul 28, 2020 16:37:08 GMT
How about them committing all redemption proceeds to repaying RYI's rather than originating new loans? Why should we be funding new lending out of money that could be used to accelerate the speed of meeting the backlog of RYI's. because if they dont then the loan book will freeze and you will be waiting years with the risk that any loan defaulting will hit your capital. Moreover, they have contracts to loan that can not be switched off overnight. Loans = provision fund increase No Loans = greater risk and % chance for capital loss for more investors.
|
|
beagle
Investor in ratesetter, funding circle, lendy (lesson learnt) and AC
Posts: 670
Likes: 322
|
Post by beagle on Jul 28, 2020 16:38:58 GMT
I've noticed that Trust Pilot for RS is posting a "great" status again but still RS site says Excellent!. I've reported it today but after reporting it in middle of July it seemed to raise itself back up with a few good reviews but over the last 2 days or so the reviews are getting very testy again. Perhaps some of the people here are reporting over there although it's not easy to tell. RS keeps saying that the Graphic where they declare the TP results on their site is automatically copied across but I'm not sure that is the case. Back in middle of July it was only when I flagged it did it suddenly change back but then they claimed it was because the TP scoring rose up again - there were some odd, almost suspect 5* 1st timer votes that I guess that pulled it back before they addressed the issue. I am not sure if they really mind what the score is at the moment or even have the time for any 'suspect' reviews.
|
|
aju
Member of DD Central
Posts: 3,496
Likes: 923
|
Post by aju on Jul 28, 2020 16:47:44 GMT
I've noticed that Trust Pilot for RS is posting a "great" status again but still RS site says Excellent!. I've reported it today but after reporting it in middle of July it seemed to raise itself back up with a few good reviews but over the last 2 days or so the reviews are getting very testy again. Perhaps some of the people here are reporting over there although it's not easy to tell. RS keeps saying that the Graphic where they declare the TP results on their site is automatically copied across but I'm not sure that is the case. Back in middle of July it was only when I flagged it did it suddenly change back but then they claimed it was because the TP scoring rose up again - there were some odd, almost suspect 5* 1st timer votes that I guess that pulled it back before they addressed the issue. I am not sure if they really mind what the score is at the moment or even have the time for any 'suspect' reviews. Well they should be in my view as it indicate a lack of quality control and my experience has been that when QA goes out the window then other things fall by the wayside.They suggested in my last correspondence - I made it a formal complaint as it wasn't the first time - that it was automatic and I wanted them to engage with the issue of maintaining correct info on the RS site especially that which encourages people to the RS site. They upheld my complaint and agreed to look into it. They have again suggested it should be automatic but have said they will look into it yet again. I'm sure it's not the case but the reviews that raised the status back up to excellent last time were not exactly kosher looking ones - very short and sweet and 1st time reviews usually about borrowing not lending. There are some bad ones about borrowing as well as more about lending. Each to their own how they see a company good or bad I suppose.
|
|
|
Post by inquiete on Jul 28, 2020 17:02:18 GMT
Just replying to posts above about RS having contracts to lend. This argument is often quoted.
It is important to note that RS are not lending. Lenders are lending. RS are matching. I don't see how RS can have any kind of obligation to lend funds on behalf of lenders. They may have provided indications to certain organisations as to target lending volumes - but they cannot have entered into any kind of binding commitment because they do not have discretionary permissions over our funds.
This is all about RS trying to look as attractive as possible to potential suitor. Again I am not sure that this should trump our interests.
|
|
|
Post by bernythedolt on Jul 28, 2020 17:21:52 GMT
Like chris1200 , I’d have (rightly or wrongly) approached this using Student's t-test on the null hypothesis, because of the low sample sizes (with the caveat that the sample variances are rather too dissimilar to provide an accurate test). The unbiased sample std dev of the weeks 8-15 data set (3.6, 4.0, 3.4, 3.5, 3.7, 3.5, 2.8, 4.3) is s 1 = 0.44 and for the weeks 16-18 data (3.6, 2.4, 2.7), s 2 = 0.62. Using the weighted average, with n 1 = 8 and n 2 = 3, I calculate the pooled std dev, s p = sqrt ( ( (n 1-1) s 12 + (n 2-1) s 22 ) / (n 1 +n 2 - 2) ) to be 0.49 Means differ by 0.7, so the t statistic is given by t = 0.7 / ( s p * sqrt (1/n 1 + 1/n 2) = 2.12 At 9 degrees of freedom, this is not significant at the 95% level (p=0.063), so I’d have concluded this isn’t quite sufficient to reject the null hypothesis! I’m looking at the data, you the standard error of their means, but why have we got such different outcomes? My degree was well over 30 years ago and I’m well rusty now, so please be gentle if I’ve made a gaffe but I’m intrigued why the two assessments lead to different outcomes, mine a borderline acceptance of H 0 but yours a definite rejection. How do we tell which is right? Could it be that looking at the data tells us more about their variance, which perhaps SEM loses sight of? Or perhaps the sample variances are too dissimilar, rendering the t-test too inaccurate to be of use. At one time I probably could have resolved this for myself, but these days I’m well out of my depth… TLDR; Sorry, you’re mostly right. I am 🦧 A bunch of small differences, which the p-value is exponentially sensitive to the t value, so that causes a big difference. Apart from one small point. #1: T distribution is correct for small numbers of items, while the standard error of the means gives the same result only in the limit of large numbers....but I carelessly misread the erfc table (2.3sigma is *not* p<0.001), and thought that it was so far over the limit that I was too lazy to do the t-distribution. #2: You used two-sided values, I used one-sided values. Two-sided is obviously correct, because if the RYIs had increased instead of decreasing we would also be having the discussion. Again, when I carelessly read p<0.001, I didn’t bother to think about which of them to use. #3: You’ve assumed that the RYI is constant *within* each period. That isn’t necessarily true. I used linear regression to de-trend and remove the gradient within each period, prior to significance testing. This decreases the s_p from 0.49 to 0.45. The linear correlation coefficient is quite high, so we are subtracting a genuine confounding factor rather than just p hacking. However, since we are removing a degree of freedom from the data, we have to use n=8 in the t-table rather than n=9, relative to a t-statistic of 2.31 Net result - I think p equals 5% Fortunately, we only have to wait another three days and this will be irrelevant That’s not too long for me to hide. No hiding necessary. Thanks for an honest and stimulating reply. I'm not sure (#3) that I've made such an assumption? I've treated the two sets as independent random sampling events collected over two discrete time periods. The exercise was to compare their sample means to say whether they could reasonably have come from the same population, or whether they could have come from separate populations, as you know. This doesn’t assume or place any bounds or constraints on the RYI data, not to my knowledge anyway (other than a reasonably comparable sample variance, 0.5 < s 1/s 2 <2 ). I think for now we're still forced to say... just about... that they are from the same population, although I do agree there is a trend developing. As you say, three days’ time should make this clearer. What really intrigues me is why you separated off the last three weeks and not the last two, which certainly do look like a new population ...we wouldn't be having this conversation!
|
|
|
Post by diversifier on Jul 28, 2020 18:01:03 GMT
I'm not sure (#3) that I've made such an assumption? I've treated the two sets as independent random sampling events collected over two discrete time periods. The exercise was to compare their sample means to say whether they could reasonably have come from the same population, or whether they could have come from separate populations, as you know. This doesn’t assume or place any bounds or constraints on the RYI data, not to my knowledge anyway (other than a reasonably comparable sample variance, 0.5 < s 1/s 2 <2 ). I think for now we're still forced to say... just about... that they are from the same population, although I do agree there is a trend developing. As you say, three days’ time should make this clearer. What really intrigues me is why you separated off the last three weeks and not the last two, which certainly do look like a new population ...we wouldn't be having this conversation! Well, I guess #3, if we are focusing on sample means rigorously, then by definition you are correct. But the underlying question I was trying to answer is “is this apparent sharp drop real? Is After different to Before?”. That’s why I want to remove the confounding effect of a (possible) slow overall reduction. If the question is “is it reducing overall over time, or is it all just random noise”, maybe it’s better to just take the entire data over both sections, stuff it into linear regression, get a correlation (R=-0.51) and work out the p value of that from t distribution. T= r sqrt(n-2)/sqrt (1-r^2) = 2.31 Which is remarkably similar to doing it the other way. I picked the last three weeks rather than two, to avoid the “Texas Sharpshooter” effect of post-selecting the target zone based on the outliers one sees. I didn’t think rigorously how large to make the group to avoid that fallacy, though, I just felt it was a trap to fall on the opposite side of.
|
|
starfished
Member of DD Central
Posts: 298
Likes: 216
|
Post by starfished on Jul 28, 2020 19:20:06 GMT
This is all about RS trying to look as attractive as possible to potential suitor. Again I am not sure that this should trum p our interests. Our or your interests? If you believe RS stats, they claim most lenders have not asked to RYI. It could be that some lenders in fact might wish that the new lending taps were even more open to hit their higher set rates... RS is playing a very difficult balancing act.
|
|
chris1200
Member of DD Central
Posts: 827
Likes: 508
|
Post by chris1200 on Jul 28, 2020 20:21:29 GMT
Our or your interests? If you believe RS stats, they claim most lenders have not asked to RYI. It could be that some lenders in fact might wish that the new lending taps were even more open to hit their higher set rates... RS is playing a very difficult balancing act. Ratesetter has 86886 investors. Between 11/3/20 and 11/7/20 they received at least 43526 RYI's. I'm not entirely sure what you're trying to prove with this stat, but it's highly possible (likely, even) for most investors to have multiple RYIs each. As this forum clearly goes to show. (Edit: And it appears that a significant number of these have been (at least partially) cancelled too, which won't be reflected in that total figure.) (Equally, many of these 'investors' may well be dormant accounts - so, in all, this stat likely doesn't tell us much at all!)
|
|
|
Post by bernythedolt on Jul 28, 2020 21:14:41 GMT
I'm not sure (#3) that I've made such an assumption? I've treated the two sets as independent random sampling events collected over two discrete time periods. The exercise was to compare their sample means to say whether they could reasonably have come from the same population, or whether they could have come from separate populations, as you know. This doesn’t assume or place any bounds or constraints on the RYI data, not to my knowledge anyway (other than a reasonably comparable sample variance, 0.5 < s 1/s 2 <2 ). I think for now we're still forced to say... just about... that they are from the same population, although I do agree there is a trend developing. As you say, three days’ time should make this clearer. What really intrigues me is why you separated off the last three weeks and not the last two, which certainly do look like a new population ...we wouldn't be having this conversation! Well, I guess #3, if we are focusing on sample means rigorously, then by definition you are correct. But the underlying question I was trying to answer is “is this apparent sharp drop real? Is After different to Before?”. That’s why I want to remove the confounding effect of a (possible) slow overall reduction. If the question is “is it reducing overall over time, or is it all just random noise”, maybe it’s better to just take the entire data over both sections, stuff it into linear regression, get a correlation (R=-0.51) and work out the p value of that from t distribution. T= r sqrt(n-2)/sqrt (1-r^2) = 2.31 Which is remarkably similar to doing it the other way. I picked the last three weeks rather than two, to avoid the “Texas Sharpshooter” effect of post-selecting the target zone based on the outliers one sees. I didn’t think rigorously how large to make the group to avoid that fallacy, though, I just felt it was a trap to fall on the opposite side of.Good on you sir! Mathematicians & scientists tend to be an honest lot, by and large, and long may it be so.
|
|
aju
Member of DD Central
Posts: 3,496
Likes: 923
|
Post by aju on Jul 28, 2020 22:59:41 GMT
I've noticed that Trust Pilot for RS is posting a "great" status again but still RS site says Excellent!. I've reported it today but after reporting it in middle of July it seemed to raise itself back up with a few good reviews but over the last 2 days or so the reviews are getting very testy again. Perhaps some of the people here are reporting over there although it's not easy to tell. RS keeps saying that the Graphic where they declare the TP results on their site is automatically copied across but I'm not sure that is the case. Back in middle of July it was only when I flagged it did it suddenly change back but then they claimed it was because the TP scoring rose up again - there were some odd, almost suspect 5* 1st timer votes that I guess that pulled it back before they addressed the issue. I think all the recent great reviews are from borrowers. Not all but you are right to a large extent, the only thing that bothered me was that there are quite a large number of reviews that look decidedly dodgy such that they tend to only have a single review and there is little text and mostly not very helpful at that too. Which Consumer magazine had a good article recently on how the fake reviews tend to take on a common look and are apparently quite easy to spot!. Worse they are becoming more prevalent too but one never can tell though as it seem companies like RS do point new members - probably borrowers - toward review sites as a norm. I personally don't tend to put reviews on sites unless I am exceptionally happy with a given product or service especially those that offer some sort of freebie I never take up, Rs offers other enticements that writing reviews I think.
|
|
aju
Member of DD Central
Posts: 3,496
Likes: 923
|
Post by aju on Jul 28, 2020 23:02:13 GMT
Ratesetter has 86886 investors. Between 11/3/20 and 11/7/20 they received at least 43526 RYI's. I'm not entirely sure what you're trying to prove with this stat, but it's highly possible (likely, even) for most investors to have multiple RYIs each. As this forum clearly goes to show. (Edit: And it appears that a significant number of these have been (at least partially) cancelled too, which won't be reflected in that total figure.) (Equally, many of these 'investors' may well be dormant accounts - so, in all, this stat likely doesn't tell us much at all!) Also it makes the assumption that the queue started from a specific position. I wonder also how many people like myself have played with it to get a feel for what works and what doesn't that will use up quite a few numbers too.
|
|