The Fed has had for many years the power to create any amount of IBDD’s (Interbank Demand Deposits). There are no reserve or reserve-ratio restrictions on the credit-creating capacity of the 12 Federal Reserve Banks.

The newly created IBDD’s can be put at the disposal of any bank in the System through the “discount window” or in the present situation the Term Auction Facility. These deposits are not only money to the recipient bank, they also become a part of the legal reserves of the System. And therein lies a limitation.

One dollar of borrowed reserves provides the same legal-economic base for the expansion of money as one dollar of non-borrowed reserves. The fact that advances have to be repaid in 28 days is immaterial. A new advance can be obtained, or the borrowing bank replaced by other borrowing banks. The importance of controlling borrowed reserves is indicated by the fact that at times nearly 84.7% of all legal reserves were borrowed (using your figure).

The Fed cannot increase the legal reserves of the System without creating the basis for a multiple expansion of the money supply (multiple in terms of the incremental reserves). The Fed can, and does, offset borrowings with open market selling operations, thus eliminating any net increase in bank legal reserves.

Therefore, with the rescue operation of the TAF’s dimension, it has offset TAF auction credit with open market selling operations of U.S. Treasury Bills
(-) 61,659 (H4.1) 1/10/08.

Great post. You are of course expert in the effect of reserve accumulation on financial conditions in these emerging market creditor economies, and impacts on growth and inflation, but I wonder though if you wouldn’t comment on the circularity here. As some analysts have pointed out, the circularity of cause and effect in reserve accumulation accounts for the above exponential chart, (and things like robust global economic growth, soaring commodity prices, and ultimately, the narrowing of the US trade deficit).

The mechanism I am referring to is that the suppression of market interest rates & spreads by recycling reserves back into the US bond market incentivized borrowing, (and concomitantly buoyed the asset prices that secured so much of it), thereby keeping the engine of growth stoked with fresh capital. The reason why this postulated simplification explains the shape of reserve accumulation is simply debt trap dynamics: finance is increasingly applied to financing costs, meaning that more and more debt is required to finance consumption growth, and this is what we have and continue to see (though the rate of debt growth is clearly decelerating). The case is even more compelling when you construct the consumer balance sheet, which manifestly shows that at least this sector (70% of the economy) is nearly totally dependent on finance for cash flow because such a large component of disposable income is spoken for.

I wonder if you are sympathetic to that simplification and if so, what might the implications be for the US and global economy of deflationary forces in the US, (namely, defaulting borrowers, slowing economic growth and an improving trade balance), and whether any event set in motion by these dynamics, (including potential central bank actions), could inspire panic in the official dollar holding community macabre economists have been speculating about for some time. Thanks…

Different parts of an investment bank aren’t allowed to talk to each other without a lawyer present. It’s not surprising that analysts have recommendations that are inconsistent from that people in the private side are doing.

DC: The flagship Goldman Alpha fund is down almost 60% in 2007 as Goldman Sachs stuffed the subprime crap from their company accounts into their customers Hedge Fund.

This isn’t what happened. I should note that I have no inside information about Goldman Alpha fund, but it is similar to other quant hedge funds that I do have publicly disclosable information about, which is part of the problem. You have so many stat hedge funds running the same strategies. Goldman Alpha has no direct CDO or subprime exposure as far as I know.

Now someone (probably not Goldman) had to liquidate some of their stocks in order to pay for subprime losses. So this caused the market to do down. Since you have all these hedge funds running the same stocks, everyone sold at the same time, which meant you have a mess on August 7, 8, 9.

This has resolved itself, and most quant hedge funds were down only slightly for Q3, and up for Q4.

This is one area where you can blame the quants, since most models for stock purchases didn’t take into account the fact that everyone else had the same model. But it is only a very small part of the total problem.

In practice, hardly anyone uses supercomputers in finance. Most systems run on clusters of standard PC’s. The reason people run clusters is that you need the risk numbers for the next trading day.

Guest: They attain a level of complexity so opaque that few people know what these things really are (if anyone does).

Simple models are usually the best because you see what is inside of them. The reason that you need massive compute power is not model complexity. Rather its because you need to run a relatively simple model against tens of thousands of instruments.

Also the simpler the model, the more computation time it takes to calculate things. Monte Carlo is dead simple to explain. I take a coin and flip it. If the coin reads heads, I assume the stock goes up. If it reads downs, it goes down. Repeat a billion times.

Analytic models are fast. I writing an equation, and calculate it. Trouble is, to reduce the problem into a single equation, I have to make dozens of assumptions and guesses.

Guest: All the while, the models were built on erroneous assumptions so simple, so fundamental, that we’re all sitting here looking at each other, saying “How could ‘smart’ people be so F’ing stupid?

They weren’t. If you run any CDO model around 2005, it’s clear that you would have a problem. So the mathematician raises the issue with his managers and one of three things happens.

1) the numbers get ignored. If some geek tells you that you should be investing in something that is bringing billions to the bank and hundreds of thousands to you personally, you can easily find a way of ignoring them if you don’t want to listen to what they have to say….

2) the numbers get listened to and the banks don’t go into some lines of business

3) the numbers get listened to and the bank comes up with a strategy to make massive amounts of money from the crash

Now you could argue that the geeks were partly responsible for the near crisis in the stock markets on August 7,8,9, but you just can’t blame “complex models” for the subprime mess. All of the mathematical models clearly showed that there was going to be a big, big problem, and anyone who says otherwise is just trying to shift blame for not listening to the quants.

US hedge funds domiciled in the carib borrowing US from US banks to buy US assets

Citi and other SIVS issuing ABCP from London to US investors to buy US assets.

CDOS domiciled offshore than buy US debt and then sell tranches to US investors.

and that is just the US — the same stuff happens with london vis a vis the eurozone and the like.

or even the yen carry trade –

when the mof was buying treausries, the flow was very direct. gross purchases matched net purchases.

now Japanese banks create structures apparently through the caribbean tho perhaps not that in turn buy US debt that are sold to retail investors and the like. of japanese banks lend to a European bank in tokyo that lends to its head office in London or elsewhere and then the yen are lent to a hedge fund that buys us assets. The net result is the same net flow, but a lot more gross flows.

I note here that recorded inflows to the US from japan have gone way way down since 04. Japanese investors probably are also taking more AUD and other kinds of currency risk, so the total rise in japanese holdings of us assets has slowed. but I doubt it has slowed as much as the us data suggests. rather, the new daisy chain operating offshore inflates the gross and obscures the ultimate source of funding for the US.

my more than 2 cents.

You’ve concluded roughly that if you strip out bank inflow/outflows and allow for financial engineering outflow/inflows, official flows are close getting close to accounting for all ‘real’ US inflows. Hence the risk.

H&K do qualify their analyses by caveats about changes in the patterns of gross flows, presumably including official flow changes.

And the global number $ 47.5 trillion outstanding of gross inflows is quite intriguing. There’s still a lot going on outside the US IIP.