In the early days of Prosper, I estimated that the cost of my idle cash (cash balance waiting to be bid, cash tied up in bids, cash tied up in loan verification, cash tied up waiting for payments to transfer to me) would be the equivalent of about 0.5% loss in interest rate. This came up in one of those old long threads where I advocated that Prosper should pay interest on cash balances.
Well I was wrong. The impact is much larger. Nearly 2%.
I know it sounds amazing.
Some of you have been following the various discussions of using IRR of cash flows into and out of prosper to judge overall portfolio performance. I originally proposed this approach because it includes the costs associated with idle cash and every other bad thing that occurs without the need to explicitly calculate them one by one. I felt that Prosper was giving folk an inappropriate measure of investment success by telling you the average interest rate on your loans, without showing you all the overhead.
When I do the IRR calculation for my portfolio I get about 11.5% . But Prosper says I have an average interest rate of 13.5% in my portfolio. The difference between those two numbers is shockingly high. I started wondering why the heck the difference was so amazingly large.
Today I took the 9 monthly statements I have received so far, and copied the numbers from the summary into a spreadsheet. One column each for:
active loan value
total account value
The ones with the * are the ones where money sits idle, not earning interest.
First thing I did was to average the columns, to produce an average cash balance, average in transit, etc. Then I took ratios of each of these items to total account value. This methodology is not precise, because of this averaging of 9 samples. I would have rather had daily data, but it is very difficult to reconstruct. Here are the results:
Average cash balance= 7.46%
Average in transit = 0.71%
Average winning bids = 3.98%
Average pending review = 1.72%
Sum of these idle cash buckets = 13.87%
Wow. On the average, 13.87% of my money has been sitting idle.
Now, my loan portfolio, according to Prosper, has an average interest rate of 13.45%. The idle cash, however, has an interest rate of zero.
So how much have I lost because this cash is idle?
13.87% x 13.45% = 1.87% !!!
To put that another way that may be clearer to some. The amount of my money that has been “active” is 1 – 13.87% = 86.13% . Therefore I’ve been earning that 13.45% average portfolio interest rate on only 86.13% of my money.
86.13% of 13.45% = 11.58% … I’ve been earning 11.58% instead of 13.45% ie losing about 1.87% due to idle cash. (This is all of course before considering the effects of defaults.)
Now this is a disappointing result for a couple of reasons. First, I’ve spent considerable effort trying to control my cash balance. I’ve made frequent deposits. During the heyday when I was ramping up my portfolio rapidly, I was making transfers every 3 days on the average. In September I slowed my bidding dramatically, but on average over the entire time I’ve been involved with Prosper, I’ve averaged one xfer every 6 days. Its not like I transferred a big pile of money in and just let it sit for months while I bid.
I suppose I could manage my cash balance more tightly. The 5 to 7 day delay makes this difficult. (Yea, I know, Prosper calls it a 4 day delay, but its 5 to 7 real days.) I have an Excel model that tells me when I should initiate a transfer. When I’ve tried to run it “tighter” I have had the experience of running out of money from time to time. Loans that meet my bidding criteria seem to come in big bursts.
Beyond managing my cash balance better, there is little I can do. I have a little bit of control over the cash tied up in bids. I could bid later in the listing cycle of course. However, I bid when I do for a reason. In the early days I bid very near the end of listings. I found that many listings terminated early, and my opportunity to bid was lost. As a result of that experience I backed up my bid time. My ongoing experience shows that was the right thing to do. There’s another recent effect that reinforces my thinking on bid timing. There are many large loans that don’t fund these days just because they don’t “catch fire”. This effect seems almost random. However, one can help them along by bidding early. Somebody’s gotta do it. There seem to be a huge number of lenders who wait to see which loans will catch fire, and then bid only after they are greater than some fraction funded.
I can’t do anything about money tied up in loan verification or payments “in transit”.
The practical realities of being a serious lender on Prosper mean that a significant fraction of one’s money is tied up in non-interest-bearing subaccounts. This results in a significant reduction in interest earned.