The Best Day to Borrow Money at Lending Club

Borrowers are getting funded pretty easily these days at Lending Club. In fact over 98% of loans have funded at Lending Club this year according to Lendstats and in my conversations with management there they say the number is even higher than that. So one could argue that every day is a great day to borrow money at Lending Club. But some days are certainly better than others.

The average amount of time to fund a loan hasn’t reduced much in the last year. Including weekends (when no loans are issued) in April 2010 it took 8.0 days for a loan to be issued and in April 2011 it took 7.7 days. Not much difference there. But if look at loans issued towards the end of the month you will see something different. Loans issued in the last two or three days of the month issue much more quickly on average.

Loans Issued in 27 Hours

Let’s take last month as an example. Some loan applications entered into Lending Club’s system on April 28 (the second last working day of the month) issued the very next day. Take a look at this loan, a $4,000 three-year credit card refinancing loan at 15.99%. It was submitted on 4/28/11 at 9:03am. It was issued the very next day, 4/29/11 at 11:58am. Or this loan, a $5,000 three-year home improvement loan at 14.79%. It was submitted at 9:16am on 4/28/11 and issued at 12:28pm on 4/29/11.

Now, that is just two loans out of the 1,563 loans issued last month. And several loans submitted on 4/28/11 took the full two weeks to fund. But if you analyze the loans from the last week of April, you will see that 4/27/11 was the best day to apply for a loan last month if you wanted your money quickly. Loans funded in an average of 6.15 days with a whopping 20 of the 52 loans submitted on that day funding in just two days.

The Very Best Day to Apply for a Loan at Lending Club

So, does that mean the fourth last day of the month is the best day to apply for a loan? Not necessarily. In March the best day of the month as far as funding average goes was March 27 where loans funded in an average of 5.41 days. So which day is the very best? That will change from month to month, but if you apply for a loan some time in the last week of the month (but not the last day) you will likely get your money more quickly, and you may get it in just a day or two.

Now, I have discussed the reason for this phenomenon before. Lending Club gets a huge influx on investor funds from their institutional investors on the last day of month. Last month it was about $4 million according to my estimates. All this money gets invested into borrower loans on the last day of the month.

Of course, with over 98% of loans funding any day is in fact a good day to apply for a loan at Lending Club. If you can get approved for a loan then you will most likely get your funding.

Subscribe
Notify of
46 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Matthew Paulson (P2P Lending News)
May. 16, 2011 1:58 pm

The statistic that 98% of loans get funded isn’t correct. 98% of loans that make it on to Lending Club’s listings might be correct, but Lending Club turns down a good 90% of borrowers before they ever make it to Lending Club’s loan listings. This just means that LC is doing a pretty good job of figuring out who should and who shouldn’t get a loan ahead of time.

Mike
Mike
May. 16, 2011 3:07 pm

Any thoughts on the recent sharp downward curve on the charts page at http://www.lendstats.com that tracks the 5 week average of the number of loans issued?

Matthew Paulson (P2P Lending News)
May. 16, 2011 3:17 pm

– Lending Club’s first week of the month is always the weakest and the last month always seems to be the strongest. As their monthly loan originations increase, it’s to be expected that that drop-off will get magnified.

Dan B
Dan B
May. 16, 2011 3:54 pm

@Matthew Paulson………While it is correct that LC does turn down approx. 90% of loan applications, & that 98%+ of listed loans fully fund, these 2 facts do not imply, as you have stated, that LC are “doing a pretty good job of figuring out who should and who shouldn’t get a loan ahead of time”.
In reality these 2 statements have very little to do with each other. 98%+ of listed loans would still fund if LC turned down 95% of loan applications or 85% of them.

Also, LC has historically always self funded a percentage of loans each month…………all the way back to when they first started. And though that number does vary from month to month, I’m sure that it does go a long way to explain the 98%+ fully funding of loans……………..as well as the highly predictable monthly loan numbers. For example this month I’ll predict that the total loan volume is going to be between $18-18.5 million.

Dan B
Dan B
May. 16, 2011 4:49 pm

@Peter………..You’re correct, I don’t think these 2 facts are related. In fact I think that LC is doing a fine & consistent job in determining which loans to approve. I just don’t think that it’s some sort of effort that is intended to insure that these loans get funded…………..because they’re going to get funded anyway, one way or another. Besides is it not true that loans will automatically issue if they get to 70% funded?

No, I’m not assuming that much of LC’s growth is due to their own funding. I’m stating that it is a FACT that they have self funded a certain percentage of loans throughout their history & continue to do so. I don’t have to speculate on a monthly volume, it is or will become public knowledge soon enough. As of Dec 31st 2010 LC has funded a total of $202.9 million in loans. Of that amount $23.8 million was self funded. That’s 11.7%. (Source Apr 2011 prospectus pg 58). I’ll leave the rest of the math to someone else.

What I am suggesting today, as I have suggested numerous times in the past is that these factors (self funding & ability to regulate the number of approved loans within a narrow band) enable them to “finesse” their numbers. There’s nothing necessarily right or wrong with this, but it is what it is. Hell I’d do the same thing if I were them.

KenL
May. 16, 2011 7:58 pm

Hi everyone, about the drop at the end of the originations chart. It’s nothing to get alarmed about, a drop like this happens every month since originations are heavy at the end of the month. This also is exacerbated by how I calculated the moving average for the current week. Since I can’t go 2 weeks into the future, the current week gets counted for 60% of the average and the previous 2 weeks get counted at 20% each. So last point of the chart (the current week) is actually calculated as 20%/20%/60%, the second week in is 20%/20%/40%/20% and the third week in and all other weeks are calculated 20%/20%/20%/20%/20%. So if the current week is slow, it really brings the tail end of the chart down. Once we get to the end of the month, that tail will shoot up again.

Lou
Lou
May. 16, 2011 8:17 pm

Hey Dan the most recent prospectus from LC states they’ve fully drawn down their credit facilities, which I interpret as they don’t have as much money to drop in at the end of the month and from now on they’ll have to rely on the institutional investors to keep their numbers going up.
@Matthew Paulson, 98% funding tells me that there’s a lot of money chasing these high returns. When the default rate drops, then I’ll say LC is getting better at figuring out who should and who shouldn’t get funded.
Lou

Dan B
Dan B
May. 16, 2011 9:33 pm

@Lou……….Where exactly does it say that? Besides that would not include the $24 million investment they received about a year ago………………so I’m not certain your point holds true.

Dan B
Dan B
May. 16, 2011 9:39 pm

I’d be surprised if this month isn’t another record breaking one. Putting aside all the charts & complicated math…………….just take the midpoint numbers for the month on the 15th, double it & add $4 million or so for the last 2 days of the month. Do that & you’ll come damn close to the actual end of month totals month in month out…………..

Lou
Lou
May. 17, 2011 8:07 am

Dan- Page 25 of the clean as filed prospectus dated 4/25/11 (page 22 if you are looking at the numbers at the bottom of the page)

“From time to time, we have relied on our credit facilities with third parties to borrow funds which we used to fund member loans on the platform ourselves to partially address the shortfall between borrower member loan requests and investor purchase commitments. We have fully drawn down these existing facilities.”

Using what’s left of their last round of investment to fund member loans depends on how much they have left, their burn rate, how soon they think they’ll be profitable, and when they think they’ll IPO. Every investment round dilutes their equity position, so, I’m not sure if they’ll trade what will be a huge gain at IPO for 9.65%.

Lou

Brian B
Brian B
May. 17, 2011 12:28 pm

I agree with Dan B about the 98% statistic not necessarily meaning LC is doing a good job screening. (if in a few years, default rate is still very low, at that point I would concede that they did a good job though)

In fact, you could just as easily use that statistic to say that the LC investors are doing a BAD job of screening which loans to fund, since all of them get funded. As average loan age matures, if default percentage goes up, I think we will be able to look back at this number and realize that “oh oops, I guess we should have exercised a bit more discretion and not automatically funded every loan that passed LC’s criteria.”

Dan B
Dan B
May. 17, 2011 2:16 pm

For an overall look at the portfolio of issued loans at LC I look at the total number of “lates”. With the ever increasing volume both in terms of total dollars as well as total loans funded one would assume that the total number of “late” loans would also track the upward trajectory. I’ve been keeping track of that number for the past 6 months or so & surprisingly that hasn’t been the case. In fact the total amount of late loans has DECLINED a little in the past 6 months.

The reason I keep track of “lates” & not defaults is because I don’t really care about defaults since they’ve already happened & “lates” are of course a leading indicator. So I’m not ready to definitively state that LC is doing a good or bad job on this or that………………….but the numbers appear to indicate that despite increasing volume, a fewer percentage of loans as well as a fewer number of loans period, are going late. I can’t see anyway that that isn’t a positive indicator.

KenL
May. 17, 2011 8:16 pm

A couple things to note. The 98% is not including listings that have been “removed”. I’m pretty sure “removed” means removed by LC because they failed screening after becoming a listing. And about 35% of listings were removed in march.

Also, going forward we will no longer be able to know how many listings expire without funding because LC has removed that from the data export. I haven’t been informed as to why, maybe they don’t want us to know this information, or maybe they are trying to make the export file smaller. If indeed they are trying to limit information, then they won’t be happy about what I’m about to do. I will soon be keeping all listings in my database and the active listings that don’t become loans I will classify them as “removed or expired”.

I have also noticed that lates seem to be down, especially over the last 6 months as Dan has noticed. Overall returns have continually improved since I have been keeping track (the last 20 months) but over the last 6 months or more they’ve really improved. I believe LC has gotten better with their screening and they may have ramped up their collections efforts too. Another big reason for so much improvement is the success of the 60 month term, which seems to get more and more popular. The 60 month term has made F & G loans quite attractive. Before the 60 month term F & G were just bad news. The 60 month term has also improved the performance of large loans as well. So now there really aren’t any significant blocks of loans that are performing poorly like there were years ago.

Bilgefisher
May. 18, 2011 7:28 am

@KenL I appreciate you holding LC accountable. The more information that tracks their loans will inevitably drive competition between LC and prosper to do better at screening loans. The old saying goes, you can’t improve what your not tracking. This will provide a much better product for both borrowers and lenders.

KenL
May. 18, 2011 7:16 pm

I really can’t think of any reason to hide that info either, I just find that data interesting, but not essential. But maybe there is something there and I should examine the data closer….. nah 🙂

Maybe the 36 month term is doing OK on the F and G loans. But back in 07 and 08, the returns were very bad on F and G’s. I was mostly avoiding them till the 60 month term came along, then I slowly warmed up to them and now I actually look for them.

Dan B
Dan B
May. 18, 2011 9:32 pm

KenL………..We’re at under 800 total F loans & under 400 G loans today at LC. I’d bet that there wasn’t more than 200 F loans & maybe as little as under 100 G loans during the ’07-’08 period that you’re referring to.

Being a stat guy I’m surprised that you’re comfortable with saying that the F & G loans were doing “very bad” in “07 & ’08……………without mentioning that the total amount of F & G loans during that period were very SMALL. So small in fact that I think it’s irresponsible to draw any conclusions from that data. I think you’re going to be hard pressed to find anyone with a stat background who will disagree with me.

KenL
May. 18, 2011 10:02 pm

A couple hundred loans with an average loss rate over 25% was enough to convince me they were bad.

Dan B
Dan B
May. 18, 2011 11:35 pm

A couple of hundred? In the F & G category each, or are we talking combined? I understand that it was enough to convince you. The question is whether it SHOULD have been enough to convince you, or anyone else that was interested in looking at it from a purely statistical angle?

KenL
May. 19, 2011 12:53 am

I was looking at F & G combined. I say the answer is YES, back in 09 the available data about F & G loans made in 07 and 08 should have convinced anyone to not bid on them. If someone knew the numbers and bid a lot on F & G listings anyway, he would have been a fool. 104 of those 282 loans have defaulted already. I’m sure that’s statistically significant.

Dan B
Dan B
May. 19, 2011 4:44 am

Your conclusions are suspect, imo. In 2009 someone looking at 2007-2008 data could have & should have interpreted that data as an aberration largely due to the tiny sample size. Contrary to what you’ve stated, if that person were to then invest in F & G notes in 2009 & on, he would have been rewarded handsomely because the loans that he was looking at in 2009 going forward have performed very well……………..so well in fact that they are now the best performing categories in Lending Club today even accounting for the bad start. Past defaults are no guarantee of future defaults. Looking at data the way you’re suggesting would put the investor in a position where he’s looking backwards & presuming that those results are a harbinger of future results…………….& that is dicey, at best.
Instead, when looking at data one should be aware of how out of whack the numbers can be when dealing with tiny sample sizes…………….& still be within normal parameters.

KenL
May. 19, 2011 9:37 am

I stand by my conclusions that I made about F and G loans and no one will be able to convince me otherwise. They performed poorly in 09 as well and the 36 month term F and G have not performed particularly well in 2010. But as I have stated before, the 60 month term have been game changers for F and G’s and they are performing great. These days there aren’t a whole lot of 36 month term F and G loans, so if I am not looking for them, I won’t be missing much anyway. There are however many more 60 month term and I do look for those.

Dan B
Dan B
May. 19, 2011 12:16 pm

@Peter………you’re talking about loans that are all done or close to the end of their 36 month term. The 36-37% default rate is very high but that’s roughly 12% per annum. What was the overall default rate for LC during that same era per annum?

LC has stated on many occasions that they were shooting for an overall default rate of under 3% per annum. Within the highest risk categories I’m fairly certain that the expectations were 9% +/- per annum defaults. With such a small sample size of 282 (across 2 categories no less), the default numbers could have ended up at anywhere between 2%- 16% defaults per annum & still be within the margin of error…………..& therefore hardly unusual.
In other words, let’s say these same numbers came back to show that only 2% of the loans defaulted per annum. Given the sample size, it would be equally irresponsible to conclude that the loans were a great investment based on the small default rate. Bottom line is that you can end up with some really skewed numbers when looking at a small sample.

There is no magic number obviously, but if I had to pick a number I’d say somewhere around 450 would be the minimum when you’re trying to draw any conclusions & even at that sample size you could be outright wrong 5% of the time.

KenL
May. 19, 2011 4:15 pm

You are right, there is no magic number, but you are wrong, the number would not be around 450. Maybe sometimes it would, but usually not. The required sample set size to prove a statistically significant difference depends on the hypothesis that is being tested. In this case that could be the difference between an acceptable default rate and the actual default rate. An acceptable default rate is subjective depending on a persons ROI goals. Instead of default rates though, I prefer to look at the difference between the actual ROI and my own personal minimum ROI goal. Back in 09 I calculated the ROI on those loans to be -9% on those loans and my personal goal back then was 10%. That’s a 19% difference. So then lets say that the hypothesis is that I want to prove is the F and G loans could not have performed 19% better than they did. Now considering that hypothesis, a sample set of 282 will more than suffice. Another viable hypothesis would be to test if those loans could have an ROI >0%. I’m sure that 282 would also have been enough to prove with 95% accuracy that those loans would never have a positive ROI.

Dan B
Dan B
May. 19, 2011 5:45 pm

Ken L………..using your own expectations as the base number in calculations is simply ridiculous. Because if one were to agree to that then you & me (or is it you & I ) & everyone else would have a different number & it’d be possible to calculate anything with any degree of certainty. My 450 would therefore be perfectly acceptable because you have no idea of what my expectations are. Do you see how silly that can get?

The main issue I have with your conclusions isn’t that F & G loans from 2007 & 2008 were bad investments. Looking back it’s obvious that they were. No one can really argue with that. The problem I have is that you seem to be using “historical data” & making generalizations going forward. Let’s say somebody signs up today at LC & decides to invest heavily in G loans. He doesn’t give a crap how the G loans from 3 or 4 yrs ago did. He wants some clue as to how the ones he’s buying today will do. What words of wisdom would you have for him then?

Dan B
Dan B
May. 19, 2011 5:57 pm

I meant impossible to calculate in the 2nd sentence

Mike
Mike
May. 19, 2011 6:38 pm

Can this argument be summarized by the ubiquitous disclaimer “Past returns are no guarantee of future results”?

KenL
May. 19, 2011 7:42 pm

Ok Dan, I agree with you on the second part. Things are changing, the economy’s changing, LC’s changing and historical data can only get you so far. For example, mortgage holders were a really bad bet back in 07 and 08, but now they are probably as good or better than renters. I still love the historical statistics however because they taught me how to think about the credit data. They showed me where many correlations are and because of them I am able to apply much more common sense than I could as a newbie. So you have to use it for what it is, but historical data surely is not the gospel. You are right, I would point to more recent historical trends and advise against looking at the 07 and 08 data if investing in G loans today.

I use whatever base number I have based on what I’m examining, I don’t choose it. I would prefer higher sample sizes, but sometimes you just don’t have that luxury, especially when we were back in 09 when the loan count was so small.

Dan B
Dan B
May. 19, 2011 11:56 pm

Ken, you & I aren’t that far apart on this. I guess my problem is that small sample sizes make me real nervous because I’ve experienced first hand how out of whack numbers can get in the short & even the intermediate term…………….. & still be a perfectly normal occurrence. Playing pro level blackjack for almost a year after college in the late 80’s gave me a great deal of respect for how wide variances can get & still be part of a normal distribution………….especially in the short term.

I am a little surprised to hear that all mortgage holders were a bad bet in 2007-08. I would have guessed that the problems would have been concentrated in the high foreclosure states & not throughout. And by the way please understand that regardless of any disagreements we may have I still think that your site is a real asset to all of us going forward.

Dan B
Dan B
May. 20, 2011 11:53 am

@Peter……..Sure, I’d agree with that but (you knew there was going to be a but) I’d be more confident if you tested the zero inquiry filter in conjunction with the other filters you normally use. Then test these filters you normally use in conjunction with 1 inquiry, 2 inquiries etc.
I have no idea if the results will come out any differently or not, but it would be the next step I’d take if I wanted to confirm the superiority of the zero inquiry filter.

Dan B
Dan B
May. 20, 2011 4:12 pm

@Peter……….It’s been quite some time since I applied for a loan & I freely admit that I know little about the mechanics involved, nor have I given much thought to any of it.
Therefore I’m perfectly ok with adhering by the collective wisdom & data that support the contention that zero inquiry is better than 1 inquiry…………..& not give it too much more thought.

But is there some plausible reason why the ROI would be higher for someone with zero inquiries rather than 1 inquiry? The reason just doesn’t jump out at me.

Xin He
Oct. 25, 2011 4:35 am

Peter when I look at # of inquiries on Nickel Steamroller, I see little correlation. Am I reading it wrong?

Craig
Craig
Nov. 18, 2011 1:14 am

Somehow they figured out I didnt need to be a borrower on the site Mat. Fact is, I have paid thousands of monthly bills(not only dollars ,but literally over 2000 notes my life), and never had a single late hit. EVER.

They are turning away a lot of good loans. Not only good loans, but GREAT loans.

My demise was that I was “self employed”.

Which brings me to this rant. I am so sick of hearing about 9% unemployement and “small business is the backbone” of this country when NO ONE will lend to us!!

Banks, Politicians — SHOVE IT!