The data. The Federal Reserve releases its quarterly flow of funds report on Thursday. Officially titled "Z.1 Financial Accounts of the United States - Flow of Funds, Balance Sheets, and Integrated Macroeconomic Accounts," the 175 page report is mind-numbing in a way only government produced documents can be. For those willing to sift through the information - or better yet, hire some economists (or archeologists) to do it for you - the report contains a wealth of high level information about the quarterly flows of financial securities and period end balance sheets of households, businesses, investors and governments.
Why it's important, but not market moving. The downside of the flow of funds report is that by the time it's published the analysis is very backward-looking. Thursday's report will update the data through 3Q 2014. The Treasury information in particular is reconstructed in part from the TICs data, so much of that information has already been made available*. Therefore any market moves that were driven over a quarter by changes in flows of, say, Treasuries, agencies, or corporate securities by a particular financial group - e.g. banks, pension funds, foreign central banks - have already taken place. Also, the data doesn't give a maturity bucketing breakdown of security buying or holdings by different investors - though it does in many cases separate Treasury bills from other Treasury securities - so the flow information it provides is "big picture".
*Sidebar: There are known and long-standing issues with the TICs data, which we won't rehash here. Government analysts have worked tirelessly over the years to improve the quality of the data, make the collection of it more timely and the data itself less subject to revision. This has in fact made the data much more relevant and useful. Despite those efforts, global reporting requirements of financial holdings are, to a great extent, voluntary. There is no system that currently exists where you can enter the cusip of a security and pull up a list of beneficial owners and how much par value they hold. "Yeah, but, the Treasury has to send payments to the bondholders, so they must be able to track who owns them." They can track the settlement details of security transactions, and the processing of the coupon or principal payments - but that tracks to the bank where the security is held, not to the owner of the account. So if the banks in Belgium hold $500 billion of Treasury securities in accounts for individuals, corporations and/or government entities all over the world, that $500 billion securities will still appear to be beneficially attributed to the country of Belgium in the TICs data because that is where the payments go. There is not (or not yet, anyway) comprehensive regulation - in the US or globally - that requires banks to turn over individual account information to the government. Of this you should be thankful.
Disclaimer to the sidebar: The above is based on my understanding of the Fedwire and Treasury payments processing systems. My understanding may be faulty or outdated, so please improve it and/or correct any mistakes, especially if you work for the Fed, the Treasury, or the FBI. (If you work for the IRS, just leave it.)
But it is useful for trend analysis and making projections. Despite the drawbacks, for our purposes the data is useful for analyzing broad financial market flows and investment trends. And one trend for at least the past decade has been that the rest of the world is the dominant financing partner of the US government. The chart below shows the amount of Treasury securities held by sector (in $ billions).
"Monetary authority" in the graph above refers to the Federal Reserve bank. Together, the US central bank and the "rest of the world" - which in the case is overwhelmingly foreign central banks - are now holding almost two-thirds of US government debt.
If we look at the quarterly flows (see the chart below), we see the same thing: that the rest of the world has most often been the dominant net buyer, while US investors - particularly households - have frequently been the dominant net sellers of Treasury securities.
Note that the quarterly flows above are seasonally adjusted annualized rates - which crudely means you need to divide by four to get the actual flows. I are a mathematician.
So what do we know. The Fed started tapering their Treasury purchases - you see that in the flows over the past two quarters as the blue bar gets smaller. Thursday's data should show Fed flows in Treasuries of zero in 3Q 2014. My guess is that you will see an increase in "rest of the world" buying, as the global economy slowed down and the dollar appreciated strongly against most other currencies in the 3Q (and continues to do so).
If I were going to make an argument for rates staying stubbornly low next year despite the Fed hiking, this is where I would start. As long as the major global economies are weak while the US is improving, there will be an overwhelmingly strong bid for Treasuries from the rest of the world.
Instead of making that argument - because you've all been inundated with "2015 Outlooks" from every sell-side and buy-side analyst on the street (and don't lie to me, I know you only read the bullet points) - what I plan to do is spend some time learning more about currencies, interest rate parity (which is apparently a great theory that seems to never actually work) and cross-asset correlations. Oh, and I will briefly update these charts on Thursday after the actual flow of funds report comes out.
No comments:
Post a Comment