by (Originally published 4 October 2008)
In a recent comment relating to the $700B banking bailout, “Bank Loans Have Not ‘Dried Up'”, on the Forbes web site, Alan Reynolds presents a strangely abbreviated set of data on total bank lending as support for his conclusion that Everyone knows that U.S. banks have virtually stopped lending, deeply slashing their loans to U.S. consumers and firms. As is so often the case, however, what everyone knows is probably not true.
I am not a strong supporter of the bailout plan, largely because I believe that it will do little, if anything, to solve the real problems in the financial system. Perhaps the best thing that could be said about the bailout plan was that it was something that could be passed quickly, and might serve to keep the financial system going until after the election, at which point real solutions might be possible.
That said, when I see a data set like that presented by Reynolds, I tend to be suspicious. I also tend to be suspicious when I see data presented that is not immediately relevant to the conclusion. In this case, that would be data regarding total outstanding loans in relation to a conclusion regarding new loans.
This is the table presented in Reynolds’ article.
| Week Ending Wednesday | Business (Commercial & Industrial) | Real Estate | Consumer | Interbank (Other Than Fed Funds) |
| Aug. 13 | 1,514.5 | 3,639.4 | 841.6 | 77.6 |
| Aug. 20 | 1,509.1 | 3,653.3 | 845.6 | 75.3 |
| Aug. 27 | 1,515.1 | 3,650.6 | 848.0 | 76.3 |
| Sept. 3 | 1,514.8 | 3,631.3 | 846.8 | 77.2 |
| Sept. 10 | 1,512.0 | 3,630.3 | 850.5 | 74.0 |
| Sept. 17 | 1,531.2 | 3,625.2 | 847.1 | 72.3 |
| Year Ago: | ||||
| Aug. 2007 | 1,311.1 | 3,498.4 | 774.0 | 82.7 |
So, what does this data tell us? It shows that the outstanding loans in these categories are more or less flat over the last six weeks, and all (except interbank loans) somewhat higher than a given date one year ago. Does this provide any evidence on the question of loans ‘drying up’?
The answer to this question is ‘no’. The data shows that lending continued over the past year — that is, that lending did not ‘dry up’ one year ago — but says nothing significant about the current situation. Indeed, the very flatness of the curve over the past six weeks, in comparison to a rising curve over the past year, is consistent with a ‘drying up’ — or at least a significant slowing — of lending. Even if all banks completely stopped new lending, the amount of total outstanding loans would not immediately plummet, as that figure indicates the total of loans made in the past (plus any new loans, and minus any loans paid, of course). Rather, the curve would first flatten (rather as appears to be the case, even given Reynolds data), and then decline.
Further, if we look at the entire data set, some interesting things appear. This is a graph of the Fed data in the categories presented by Reynolds, going back to 1973.
What this shows is a generally rising lending total in the four categories referenced by Reynolds, with a few dips. A number of things could be said about this graph, including the staggering rise and rise of real estate lending, but perhaps most relevant is the very end of the graph, which displays a slight change. This change is much more visible if we include just the data from 2000 to the present.
That is, looking at all categories except consumer loans, the curves have become either flat or negative, and even the consumer curve is very nearly flat. The same is shown by the data jusf from August 2007 to the present.
Of course, none of this data is conclusive of anything. It is indicative of a lending slowdown, which almost certainly is occurring, but does not demonstrate a ‘drying up’. But what is most important is that it very much does not provide any sort of evidence against a ‘drying up’ argument.
