50 replies to this topic

[Ledney]

Not sure what this change in Ratesetters means. I assume some other company is now passing on loan requests. Good or Bad?

[rudry2677]

It's almost a Political sound-bite informing everyone that all is ok.  Is it?   More clarity please RS.

[MRC_London]

 rudry2677, on 16 August 2012 - 21:25, said:

More clarity please

Have you seen their blog post  ?

[alexp2000]

 MRC_London, on 16 August 2012 - 21:49, said:

Have you seen their blog post  ?

Yes, and it doesn't help!
Not convinced I'm afraid. They remain less than transparent with regards packaging up their monthly loans into longer term loan contracts. On the face of it this new change appears to be more obfuscation of the exact goings on - making them more and more like a bank.

[propman]

 alexp2000, on 16 August 2012 - 22:58, said:

Not convinced I'm afraid. They remain less than transparent with regards packaging up their monthly loans into longer term loan contracts. On the face of it this new change appears to be more obfuscation of the exact goings on - making them more and more like a bank.

My reading was that their credit criterias hadn't changed and the loan term differences were not significant, so no-one should see any difference except perhaps an increase in lending. Any difference is being kept quiet and any change has risks, but compared to the risk on their 60 month oanbook blowing up I'm not going to give it much thought. neither is anyone else by the look of things as funds for lending hae continued to grow and rates fall.

Re the packaging of loans, I found Rhydian very open on the subject. Basically at each month end any withdrawals (including, I surmise, those lenders who specify a minimum rate above the "current rate") are funded first from the monthly payment (or provision Fund if no payment is made on time), then from funds from the market including the funds relent at higher rates than the current rate. This will work fine until their is a run on RS. If there are not sufficient funds, the borrowers will be asked to choose between applying to roll into a longer market (if accepted) or asked to repay in full. If they fail to do so then i am assured the funds will not be taken from the provision, instead the lenders will have to wait until the loan is repaid on its normal schedule or funds are avialable to refinance. I did not ask whether they would "refinance" mid-month if funds became available. In addition, after a year, the whole balance is refinanced on the monthly market as the remainder of the funds have been borrowed on the 12 month market. Of course these will not start to crystalise until next February.

I agree that the risk of having to wait beyond the month is not made plain to lenders, but I think the bigger issue is for borrowers whose rate could rocket if the only funds were oportunistic funds who knew that any offer would be accepted on the borrowers behalf! I am steering well clear of the monthly and annual markets as the returns are very low when you consider these risks and I do not expect to need my P2P money any time soon.

To be fair, I think they are more open than most companies, in my experience Zopa are no more so initially although the forum does hold them to account afterwards. Of course their may be more interplay with "tester" lenders at Zopa, but I can'[t comment on that.


As always JMHO
- PM

[Ledney]

Ratesetters are also now re-assigning 'Market Rate' loans every day. I think this is driving rates down as I assume all the money being lent at 'Market Rate' is moved in a chunk to the (nearly) lowest rate every day.

This is what was predicted would happen at Zopa, if a follow the market rate facility was initiated.

[elljay]

 Ledney, on 07 October 2012 - 08:05, said:

Ratesetters are also now re-assigning 'Market Rate' loans every day. I think this is driving rates down as I assume all the money being lent at 'Market Rate' is moved in a chunk to the (nearly) lowest rate every day.

This is what was predicted would happen at Zopa, if a follow the market rate facility was initiated.

It's reasonably clear on the graphs on the Monthly and Yearly markets, but the 3 year and 5 year seemed to be pretty stable.  Since the first of October all the rates seem to be heading toward the floor though!  I guess some of it is the loans automatically following the lowest rate down, but there seems to be quite a bit of money still at each of the higher rates, so it can't all be that.

[propman]

 elljay, on 07 October 2012 - 08:12, said:

It's reasonably clear on the graphs on the Monthly and Yearly markets, but the 3 year and 5 year seemed to be pretty stable.  Since the first of October all the rates seem to be heading toward the floor though!  I guess some of it is the loans automatically following the lowest rate down, but there seems to be quite a bit of money still at each of the higher rates, so it can't all be that.

We have seen similar reductions in rates before where a significant number of lenders race to the bottom, the new method will just exasperate it. I suspect that the build up at rates somewhat above the current market is a combination of people happy to stick at their "minimum" rate, those who selected a pre-determined minimum for relending that is now off market and people who think this is temporary waiting out their hoped for increase in rates.

The Longer market (called 60 months, but includes some 48 month loans!) has now shown a modest improvement from a low of around 5.8% due to strong demand, but still someway off the 6.8 & 7% rates where the largest amount of money is on offer (assuming rates recover to the rates achieved at the end of last month, this last happened in June).

- PM

[PoohBah]

 propman, on 08 October 2012 - 14:44, said:

... the new method will just exasperate it. ...

It will probably exacerbate it, while undoubtedly exasperating the lenders!

Mrs Malaprop is alive and well and living in this forum.

[sl75]

 Ledney, on 07 October 2012 - 08:05, said:

Ratesetters are also now re-assigning 'Market Rate' loans every day. I think this is driving rates down as I assume all the money being lent at 'Market Rate' is moved in a chunk to the (nearly) lowest rate every day.

The same effect would ALSO drive rates up in cases where the daily demand exceeded the daily supply of funds.... rather than "driving down" as such, it's "driving towards the point where supply and demand are balanced".

It's only when lenders manually choosing their rate decide to undercut the current lowest offer because they believe it to be too slow when waiting at the end of the "current" lowest rate, and those undercutting offers aren't "instantly" matched by short-term demand that rates would get driven down.

(or to put it another way - if, at the present rates, you prefer to shave another 0.1% off than either to wait or to remove your money from the system entirely, and act on that preference then YOU are part of the force that is driving rates down.)

Edited by sl75, 08 October 2012 - 15:53.

[Ledney]

True, but the temptation to go just 0.1% down, particularly for new lenders, is great. I don't know how much of a chunk the Market following money is, or whether it is itself discounted when assessing the new Market rate offer. If there were a lot more borrowers perhaps it would drive rates up, maybe it's just a bit low on borrowers at present. Have to wait and see, like the new Zopa system.

[sl75]

 Ledney, on 08 October 2012 - 16:33, said:

True, but the temptation to go just 0.1% down, particularly for new lenders, is great.

Sure, but in the presence of strong demand, any offer reduced by 0.1% will soon be wiped out... and the "Market Rate" is set "at the level of the Lowest Lender Order, after discounting a small percentage of all the offers". I forget if the actual percentage was specified somewhere else, but it's effectively set at a low percentile rather than the actual lowest offer.

So market rate falling implies that the number of lenders who have undercut the market rate by at least 0.1% has exceeded the subsequent demand for funds from borrowers, and market rate rising implies that sufficient borrowers have been taking out loans to clear out all of that day's funds at the previous market rate (whether that's because there are hardly any lenders still lending at that market rate, or that there've been an unusually large number of borrowers).

For lenders who are still prepared to lend at the market rate, their system simply adjusts more quickly to the new balance (a typical relatively passive Zopa lender might take several weeks or even months to realise the rest of the market has adjusted, and adjust their own rates accordingly. However, a passive RateSetter lender will react about as quickly as a more active lender).

[propman]

I think the issue is that when supply at the "market rate" significantly exceeds demand, then there is a significant amount of money left at the lowest rate (ignoring small offers made to undercut the market whn these are still on the market as the new funds are added). I am not sure whether there is a manual change of rates / new crop of fresh money required first to bring the rate lower, or whether the "market" rate is set to ensure rapid lending and so automatically lowers the rate on the grounds that otherwise there will be a delay before funds are relent. However, in practice there is a quantum of money beyond which lenders will assume that it is worth losing 0.1% to queue jump. I suspect tha the higher demand not be fully factored in, so for many lenders this may be set based on significantly lower lending volumes. In any event, the repayment run was appearing to lower the rate regularly. By adding previously auto lent money at the lower rate, the larger lending runs continue to lower rates through several days.

JMHO

- PM

[Ledney]

What I was trying to say, I think.

[propman]

 Ledney, on 08 October 2012 - 19:58, said:

What I was trying to say, I think.

Sorry, I should have said that I was agreeing with you and addressing what I saw as the limitations on Sl75's reasoning in the current market.

On another point, today the current achievable rate on RS's longer market (they call it 60 month but make 48 month loans from it as well) has sunk to the Zopa rate assuming predicted bad debt, 1% fee on Zopa AND a 40% tax rate and providing no discount for any greater risk of the RS platform, lack of emergency access mid-term or any risk that the provision fund might be insufficient. The impact (assuming that a Zopa A* borrower would only be charged the minimum 0.5% credit fee on RS) is that Zopa is only cheaper for borrowings of £5-8k.

Strangely there is now only a 0.5% rate differential on RS between 36 & 60 month markets for lenders against a 1.2% difference on Zopa! The only reason I can think is that RS has a higher proportio of larger lenders (>£11k average per borrower against Zopa's <£3k) and these are more likely to have funds available for the longer terms.

- PM

[sl75]

 propman, on 15 October 2012 - 15:39, said:

Sorry, I should have said that I was agreeing with you and addressing what I saw as the limitations on Sl75's reasoning in the current market.

Your own argument seemed to be predicated on their being "a significant amount of money left at the lowest rate".
Significant compared to what?

I'd tend to think a clearer view of the actual market would be created by using "£xxxxx of matches in last 24 hours" (or even better ".... in last 7 days") rather than "NNN matches in last 24 hours", allowing lenders to put the length of the various queues into proper context, and avoid unnecessarily undercutting a rate where the queue is likely to be fully turned over within a very short space of time, but just looks quite long when there's no context.

However, with a little digging, a reasonable estimate can be derived. I see right now that in the "monthly access" market there were 545 matches in the last 24 hours. The full market data gives both the amount of money on offer and also the number of orders that represents, suggesting an average order size in the region of £1k. Assuming an average of between 1 and 2 matches per "order", and that the current market is representative of the typical order size, that would suggest that something in the region of £300k to £500k had been matched in the last 24 hours... much of it being from the loans which rolled over this morning as the monthly payments for any loan with a payment due over the weekend were processed.

In context, that £138.7k queue with 161 orders at 3.0% doesn't look so huge after all... and even though we'd only expect about a third of the volume overnight tonight (loans with a payment day of 16th rolling over) compared to last night (loans with payment days of 13th, 14th or 15th), it should represent a similar order or magnitude to the overnight matches, suggesting that, unless there are enough offers at 2.9% to make that "the market rate" tonight, the current orders in the 3.0% queue could be [almost] completely replaced by new orders from automatic re-investment instructions overnight.

[propman]

 sl75, on 15 October 2012 - 18:12, said:

Your own argument seemed to be predicated on their being "a significant amount of money left at the lowest rate".
Significant compared to what?

I'd tend to think a clearer view of the actual market would be created by using "£xxxxx of matches in last 24 hours" (or even better ".... in last 7 days") rather than "NNN matches in last 24 hours", allowing lenders to put the length of the various queues into proper context, and avoid unnecessarily undercutting a rate where the queue is likely to be fully turned over within a very short space of time, but just looks quite long when there's no context.

I agree that it is far from clear. Some offers can be £50k so quite a few loans are made from a single match. That said, the site quotes an average of 23 borrowers to each lender with the average lender having £11,727 deposited and about 1/14th of the money not lent out, that suggests about the £475 average offer or 8-9 offers per loan. I doubt that RS see it as in their interest to reduce the under cutting. The total expected matches (across all 4 lender queues) is quoted, currently £158k in next 24 hours. Some 30-40 average loans. I don't invest in the monthly market, but £60k at a single rate on the other markets would seem unlikely to be all matched in the day as there is daily additional deposits on top of the repayments. Also the average loan term of 28 months suggests a majority of the loans are from the monthly / yearly funds making movement in the 36 & 60 month markets slower.

 sl75, on 15 October 2012 - 18:12, said:

However, with a little digging, a reasonable estimate can be derived. I see right now that in the "monthly access" market there were 545 matches in the last 24 hours. The full market data gives both the amount of money on offer and also the number of orders that represents, suggesting an average order size in the region of £1k. Assuming an average of between 1 and 2 matches per "order", and that the current market is representative of the typical order size, that would suggest that something in the region of £300k to £500k had been matched in the last 24 hours... much of it being from the loans which rolled over this morning as the monthly payments for any loan with a payment due over the weekend were processed.

In context, that £138.7k queue with 161 orders at 3.0% doesn't look so huge after all... and even though we'd only expect about a third of the volume overnight tonight (loans with a payment day of 16th rolling over) compared to last night (loans with payment days of 13th, 14th or 15th), it should represent a similar order or magnitude to the overnight matches, suggesting that, unless there are enough offers at 2.9% to make that "the market rate" tonight, the current orders in the 3.0% queue could be [almost] completely replaced by new orders from automatic re-investment instructions overnight.

The rollover makes the monthly payments misleading. I have not followed the queues, but suspect that most of the rollovers are covered by automatic rematching the money that is being rolled over, I suspect that this is never shown in the queues as it happens as soon as the money is released. Anyway you do not usually find the rate lifting appreciably on a Monday as would be expected if a high proportion was new money.

JMHO - PM

[sl75]

 propman, on 16 October 2012 - 08:31, said:

The rollover makes the monthly payments misleading. I have not followed the queues, but suspect that most of the rollovers are covered by automatic rematching the money that is being rolled over, I suspect that this is never shown in the queues as it happens as soon as the money is released.

I think you possibly misunderstand the operation of the rolling monthly market - either that or I do.

I see no reason for the automatic rematching of money that is being rolled over not to use the same queues as the rest of the money - why do you suppose otherwise?

There's a huge spike that occurs in the number of matches during the wee small hours. It used to be visible on elljay's charts before these stopped being updated for RS (it's a bit less clearly visible on the 6 months ago chart, as that has been pruned down to a 12 hour average). That would occur from all of the existing lenders being giving their money back, instantly re-investing it, and all of the borrowers being re-matched to a new lender (in whatever order those various operations actually occur!)... and if the amount that rolls is significantly larger than the amount already on offer at market rate, most lenders would get rematched during that process. An optimised process need only take a few seconds to work through rematching all the rolling loans due on that day, so most lenders would indeed get rematched as soon as the money is released, but to a different borrower, having entered one end of the queue and exited through the other.

The process would generally run a lot faster that Zopa's payment run, because a typical match would be much larger, and electronically handling movements of £1000+ takes no more processing power than electronically handling movements of £10... and also because they have rather fewer outstanding loans (and I'd hazard a guess they probably have a more efficient back end).

[propman]

Rereading I now agree with you. I had assumed that the "contract" was for the borrowers term with a right to replace the lender on rollover days. I cannot access the contract for the only one I tried as an experiment. However, this should not effect the point I was making as it will frequently (possibly usually) only allow rates at the "market rate" to be matched. I realised that the matches included rolled over funds, hence most of the "Queue" will not be available to participate in these rolling over matches. This may become less of an issue once the anniversary of the 12 month market passes as presumably this will mean that the rollovers should usually exceed the money rolled over unless most of the matches to the monthly market are for 6 or 12 months and therefore are not matched to the yearly funds.

In any event, the point is that if there is enough money left at the "market rate" after the rollover that borrowers worry whetehr their funds will be matched, then there is likely to be enough funds offered at lower rates for the market rate to fall. While I agree that the matching in the monthly markets may make the threshold high enough to stop this being an issue, in the other markets where it is only new loans that are matched, the threshold will be lower and hence my conclusion more likely to be valid.

- PM

[propman]

Apologies for the double post.

Interestingly the rates recently increased (1% on longer market and 0.5% on 36 month), I believe that this is the most rapid increase since they became properly established. This has happenned due to a 1/3 reduction in the "lender balances" and halving of funds on offer in these 2 markets. So contrary to what I said above, like the proposed issue raised with Zopa's change, the impact has been to reduce the resilience as well as the absolute rates thus making them more volatile. From RS's point of view, I suppose the net impact is likely to have increased the number of loans made. Certainly there was a rush of loans when the rate dipped below Zopa's for sub £5k loans. However, if I was Rhydian, if this were to happen regularly, I would be concerned that I might be turning away potential customers who applied based on one rate but were then offered a much higher rate (as the matching is on acceptance rather than when the application proceeds).

JMHO

- PM