tag:blogger.com,1999:blog-2499715909956774229.post817246199359963998..comments2024-03-22T22:37:02.639-07:00Comments on Stephen Williamson: New Monetarist Economics: Fed Balance Sheet NewsStephen Williamsonhttp://www.blogger.com/profile/01434465858419028592noreply@blogger.comBlogger27125tag:blogger.com,1999:blog-2499715909956774229.post-51592983476419576592018-07-09T15:25:56.451-07:002018-07-09T15:25:56.451-07:00Thanks for the information. It's really useful...Thanks for the information. It's really useful for everyone to know the institutional details of these markets. There's a view among some of the policymakers that these details aren't important for policy, but I don't think that's right.Stephen Williamsonhttps://www.blogger.com/profile/01434465858419028592noreply@blogger.comtag:blogger.com,1999:blog-2499715909956774229.post-75143948384649502452018-07-09T14:00:49.996-07:002018-07-09T14:00:49.996-07:00Thanks for your reply. One small point of clarifi...Thanks for your reply. One small point of clarification specific to SOFR as a potential rate benchmark. The bilateral repo segment included as BGCR steps up to SOFR includes the activity of many smaller firms as collateral providing counterparties. If liquidity transmission were clogged in this channel, then by assumption dealer-facing OMOs would not be effective in combating a widening bimodality. The median would drift higher. It's hard to say whether this clogging scenario is likely to crop up over the coming years, but it's something potentially worth considering when devising an all-weather approach.Chrishttps://www.blogger.com/profile/02252134161181976682noreply@blogger.comtag:blogger.com,1999:blog-2499715909956774229.post-47341574501392703192018-07-08T12:22:17.732-07:002018-07-08T12:22:17.732-07:00"However, I do wonder if hitching the policy ..."However, I do wonder if hitching the policy target to the big repo pools puts the Fed on the hook, in some sense, for too large a universe of funding, or market dynamics that lie outside their direct control."<br /><br />If the Fed can't control the overnight repo rate, they have a big problem. The approach before the financial crisis was to intervene in the repo market in order to influence what is going on in the fed funds market, and that seemed to work, more or less. The model the New York Fed people had in their heads for that implementation approach was a loanable funds model - fed funds rate is determined by by demand and supply of reserves, and the idea was to have some notion of the demand curve for reserves on a given day, and to intervene in the repo market so that the market would clear at the target fed funds rate. But what they were actually doing was moving the repo rate around day to day so as to hit the fed funds rate target. It's a lot easier just to target the repo rate as you can actually see it - it's the rate on the transactions you're actually engaged in.<br /><br />"Incidentally a switch back to a point target could be useful in that regard. The day-to-day volatility of SOFR is not much higher than fed funds was in the late 1990s."<br /><br />Yes, I agree.Stephen Williamsonhttps://www.blogger.com/profile/01434465858419028592noreply@blogger.comtag:blogger.com,1999:blog-2499715909956774229.post-58420655106016590822018-07-07T11:02:00.697-07:002018-07-07T11:02:00.697-07:00According to the information found at the URL refe...According to the information found at the URL referenced, the distribution of money by denomination is dominated by $100 bills by value. The next largest in value is $20 bills. Annual compound growth rates, by volume, by denomination since 2010 (as determined by OLS regression) are as follows (r-squared in brackets): $100 - 8.4%/yr (0.9939); $50 - 3.69%/yr (0.9272); $20 - 4.96%/yr (0.9865); $10 - 2.30%/yr (0.9253); $5 - 3.67%/yr (0.9678); and, $1 - 3.23%/yr (0.9988).<br /><br />If one assumes that the demand for $100 bills is primarily related to the drug trade and that there are three intermediaries between the domestic user and the foreign producer, and each intermediary skims 10% of the take he receives (a Markov process assumed), 27% of the users' $100 bills withdrawn from domestic financial institutions remain in the domestic economy and the balance of 73% makes its way into the off-shore economies. <br /><br />For year-end 2017, that model would put the value of $100 bills in domestic circulation at roughly $340 billion compared to roughly $320 billion of value in all other denominations. The weighted average growth rate of the domestic portion of currency in circulation, allowing 8.4%/yr ACGR for $100 bills and 3.5%/yr ACGR for all other lesser denominations in the aggregate, approximates 6%/yr ACGR over the past 7 years. Though not as impressive as an 8%/yr ACGR, 6%/yr is still impressive. Of course, this picture is founded on a series of untested assumptions (sand?) and as a result it can only be considered to be a thought-experiment (gendankenexperiment).<br /><br />The value weighted-average ACGR for denominations of $1 to $50 for the period commencing 2011-01-01 through 2017-12-31 is 4.33%/yr, leading to a weighted average ACGR of 6.33%/yr, when the figures are pushed through a spreadsheet calculator. <br /><br />Even at 4% per year, the expansion of currency in circulation is double the Federal Reserve's target for inflation (howsoever one measures inflation). Then, again, one must consider that the foreign-held component of currency in circulation must eventually feed back through commodity prices (whether the commodities be resource commodities or engineered commodities) as a result of the ever greater amounts of dollars in circulation in off-shore economies where the commodities originate from. <br /><br />If Milton Friedman was correct in stating that inflation is almost everywhere a monetary phenomenon, then we are probably looking at a day of reckoning such as we once faced in 1981-2. On the other hand, if the New Monetarist view of inflation based on Fisher's postulate holds, then we can avoid that day of reckoning by keeping nominal interest rates low (close to the ZLB) forever. We have, perhaps, another ten years to find out. (Après nous, le déluge?)<br />Old Eagle Eyehttps://www.blogger.com/profile/05270080708077871311noreply@blogger.comtag:blogger.com,1999:blog-2499715909956774229.post-83822914878271470252018-07-07T07:45:40.885-07:002018-07-07T07:45:40.885-07:00Stephen, thanks for this stimulating discussion.
...Stephen, thanks for this stimulating discussion. <br /><br />I think you are right that the FOMC needs to consider switching to a repo interest rate benchmark. Going back to the early days of LSAP, Fed officials have expressed concern that structural changes in the fed funds market have decreased its relevance to financial firms and the economy. At around $80B per day, the fed funds market is dwarfed by the repo pools tracked by FRBNY’s TGCR and SOFR indexes ($350B and $700B per day, respectively). The Fed has shown a strong preference for secured lending since the crisis so a move to repo would be helpful there. Presumably it would also accelerate adoption of SOFR as a LIBOR replacement, a stated goal of Powell's.<br /><br />However, I do wonder if hitching the policy target to the big repo pools puts the Fed on the hook, in some sense, for too large a universe of funding, or market dynamics that lie outside their direct control. As with LIBOR, it’s reasonable to expect that spreads between IOER and repo benchmarks will widen and narrow as money market conditions change, and presumably the Fed does not want to be always explaining small movements or temporary violations of target ranges. <br /><br />Incidentally a switch back to a point target could be useful in that regard. The day-to-day volatility of SOFR is not much higher than fed funds was in the late 1990s.<br />Chrishttps://www.blogger.com/profile/02252134161181976682noreply@blogger.comtag:blogger.com,1999:blog-2499715909956774229.post-79685898524604928712018-07-06T15:15:03.664-07:002018-07-06T15:15:03.664-07:00no problem.no problem.Stephen Williamsonhttps://www.blogger.com/profile/01434465858419028592noreply@blogger.comtag:blogger.com,1999:blog-2499715909956774229.post-89778102433159605302018-07-06T15:14:10.743-07:002018-07-06T15:14:10.743-07:00Thanks.Thanks.Stephen Williamsonhttps://www.blogger.com/profile/01434465858419028592noreply@blogger.comtag:blogger.com,1999:blog-2499715909956774229.post-69735384902654837842018-07-06T15:12:26.496-07:002018-07-06T15:12:26.496-07:00Suppose they had set the ON-RRP rate equal to IOER...Suppose they had set the ON-RRP rate equal to IOER. I think that would have mostly undone the scarcity (and undone the policy too I think), but there would have been a lot of takeup on the ON-RRP facility - likely most of the reserves would have turned into ON-RRPs. But we wouldn't think that did any harm, I think, as the ON-RRPs are just reserves by another name.Stephen Williamsonhttps://www.blogger.com/profile/01434465858419028592noreply@blogger.comtag:blogger.com,1999:blog-2499715909956774229.post-33431738348948147162018-07-06T15:06:50.441-07:002018-07-06T15:06:50.441-07:00See my next post.See my next post.Stephen Williamsonhttps://www.blogger.com/profile/01434465858419028592noreply@blogger.comtag:blogger.com,1999:blog-2499715909956774229.post-84631682116024449562018-07-06T13:22:52.734-07:002018-07-06T13:22:52.734-07:00Thanks Steve! A very interesting discussion. Thanks Steve! A very interesting discussion. David Beckworthhttps://www.blogger.com/profile/04577612979801459194noreply@blogger.comtag:blogger.com,1999:blog-2499715909956774229.post-55609629932589655752018-07-06T13:21:59.182-07:002018-07-06T13:21:59.182-07:00Thanks Steve. Yes, I am surprised on the first poi...Thanks Steve. Yes, I am surprised on the first point as well (and need to go revise a paper based on this observation). <br /><br />On the second point, I agree with you that it's best to get rid of the spreads. I was just wondering if the Fed accidentally stumbled into a way of encumbering the reserves given that is what they initially wanted to do with IOER. Sometimes it is better to be lucky than smart. David Beckworthhttps://www.blogger.com/profile/04577612979801459194noreply@blogger.comtag:blogger.com,1999:blog-2499715909956774229.post-26962438639982242662018-07-06T12:02:29.422-07:002018-07-06T12:02:29.422-07:00One of the more interesting panel discussions I sa...One of the more interesting panel discussions I saw involved a bunch of debt managers (for the US, UK, German, Spanish, and Danish governments) talking about what they did. Spanish debt manager talked about responding to complaints by market participants about a shortage of long government debt caused by the ECB, which was buying it all.Stephen Williamsonhttps://www.blogger.com/profile/01434465858419028592noreply@blogger.comtag:blogger.com,1999:blog-2499715909956774229.post-46577835116495520592018-07-06T11:58:06.865-07:002018-07-06T11:58:06.865-07:00I'm not sure where to find numbers on this. My...I'm not sure where to find numbers on this. My impression is that Treasuries of any maturity work in a repo transaction. Long maturities are associated with larger haircuts, but as haircuts go, I think the differences aren't that big. I think on-the-run is preferred to off-the-run, and on-the-run is going to be lean toward bills as most of what gets rolled over each month is bills. That said, there's fungibility, and I don't think the Treasury's behavior is invariant to what the Fed is up to. The Treasury has to float a given quantity of new debt every month, and it worries about satisfying the demand for particular maturities. If the Fed starts buying more long-maturity bonds, the Treasury would tend to adjust by issuing more bonds and less bills. Then when reinvestment stops, the Fed takes pressure off the bond market, and the Treasury issues more bills.Stephen Williamsonhttps://www.blogger.com/profile/01434465858419028592noreply@blogger.comtag:blogger.com,1999:blog-2499715909956774229.post-16143109276326547832018-07-06T11:50:34.268-07:002018-07-06T11:50:34.268-07:00"First, in general, does this observation mea..."First, in general, does this observation mean that given sufficient collateral for overnight lenders, the limited access to the Fed's balance sheet should not be an important financial friction? That is, arbitrage should reduce spreads among short-term rates to very small values even if only a few financial firms have access to the Fed's balance sheet?"<br /><br />Yes, I think so. I'm surprised by this, as it goes against what I think we knew.<br /><br />"Second, maybe the gumming up of overnight markets worked inadvertently to help the Fed. Here's how: Fed LSAPs create the IOER spread over other short-term interest rates. This causes the relative return on excess reserves to go up and raises their demand by banks. This heightened demand effectively "sterilizes" or encumbered the large stock of excess reserves."<br /><br />This is more complicated than it might seem, as it involves how nominal interest rates (and which ones) feed through to inflation. I think it's a misconception to think that you want to somehow "encumber" the reserves to make sure they don't leak out and cause inflation. Any number of bad policy proposals stem from that notion (remember term deposits - that was about encumbering reserves). I think it's best to think of these interest differentials as reflecting inefficiency, and that it's best to get rid of them.<br />Stephen Williamsonhttps://www.blogger.com/profile/01434465858419028592noreply@blogger.comtag:blogger.com,1999:blog-2499715909956774229.post-57700336977914452542018-07-06T09:56:00.947-07:002018-07-06T09:56:00.947-07:00Steve,
Been thinking a bit more about your post a...Steve,<br /><br />Been thinking a bit more about your post and have a follow-up question. The Fed has not held treasury bills for some time, so the reinvestments of principal it was previously doing was in medium-to-longer term treasury securities. Thus, its flow of asset purchases was not on treasury bills. But aren't treasury bills the most widely used with repos? David Beckworthhttps://www.blogger.com/profile/04577612979801459194noreply@blogger.comtag:blogger.com,1999:blog-2499715909956774229.post-19118573890395836572018-07-06T09:33:19.944-07:002018-07-06T09:33:19.944-07:00Fred has discontinued some ON-RRP time series. How...Fred has discontinued some ON-RRP time series. However, by using New York Fed's data set, I created a chart including the value of ON-RRP outstanding and the effective fed funds rate: <br /><br />https://plus.google.com/photos/photo/107299763175419710968/6575140811952527314<br /><br />Could be useful. Thanks for the post. Anonymoushttps://www.blogger.com/profile/17974910528536453844noreply@blogger.comtag:blogger.com,1999:blog-2499715909956774229.post-65193158906564069762018-07-06T08:05:38.913-07:002018-07-06T08:05:38.913-07:00Steve, very interesting post.
I have several que...Steve, very interesting post. <br /><br />I have several questions regarding your general conclusion nicely summarized by this statement " The IOER/fed funds rate margin was not caused primarily by "balance sheet costs," but by lack of good alternatives to the fed funds market for overnight lenders." <br /><br />First, in general, does this observation mean that given sufficient collateral for overnight lenders, the limited access to the Fed's balance sheet should not be an important financial friction? That is, arbitrage should reduce spreads among short-term rates to very small values even if only a few financial firms have access to the Fed's balance sheet? <br /><br />Second, maybe the gumming up of overnight markets worked inadvertently to help the Fed. Here's how: Fed LSAPs create the IOER spread over other short-term interest rates. This causes the relative return on excess reserves to go up and raises their demand by banks. This heightened demand effectively "sterilizes" or encumbered the large stock of excess reserves. <br /><br />David Beckworthhttps://www.blogger.com/profile/04577612979801459194noreply@blogger.comtag:blogger.com,1999:blog-2499715909956774229.post-82282236463807207862018-07-06T06:28:33.444-07:002018-07-06T06:28:33.444-07:00It's useful to look at currency outstanding by...It's useful to look at currency outstanding by denomination. A lot of that growth is in the stock of $100 bills: https://www.federalreserve.gov/paymentsystems/coin_data.htmStephen Williamsonhttps://www.blogger.com/profile/01434465858419028592noreply@blogger.comtag:blogger.com,1999:blog-2499715909956774229.post-27543214622861119842018-07-06T06:23:43.244-07:002018-07-06T06:23:43.244-07:00"...in the latest minutes the Fed seems to su..."...in the latest minutes the Fed seems to suggest that the rise in o/n repo rate is associated with a substantial increase in the net supply of Treasury bills."<br /><br />Yes, I saw that in the minutes. Treasury issuance was pretty strong, I think in April and May, but it was down in June. Generally, issuance is quite variable, and it's never shown up before in the overnight rates in this way, so I don't buy it.Stephen Williamsonhttps://www.blogger.com/profile/01434465858419028592noreply@blogger.comtag:blogger.com,1999:blog-2499715909956774229.post-88676682321287080642018-07-05T21:05:45.021-07:002018-07-05T21:05:45.021-07:00Since 2009, currency in circulation has grown at a...Since 2009, currency in circulation has grown at an average annual compound growth rate of 8 percent. This is well in excess of the posted rate of inflation. Much of this currency is undoubtedly moved overseas. But, even so, currency in circulation has doubled in 9 years. Is there any reason to be concerned that inflation will pick up, as it did in the 1970s as a result of this expansion in currency in circulation?<br /><br />One other point worth mentioning: The trend in growth in currency in circulation shows no breaks or reversals in that 9 year period, and it gives every indication that the trend will continue at the present average rate. What effect, if any, might this have on future commodity price levels (with the term 'commodity' interpreted broadly)?Old Eagle Eyehttps://www.blogger.com/profile/05270080708077871311noreply@blogger.comtag:blogger.com,1999:blog-2499715909956774229.post-35767592675489676272018-07-05T18:50:47.292-07:002018-07-05T18:50:47.292-07:00Stephen - insightful post. Hoover Institution had ...Stephen - insightful post. Hoover Institution had a related conference on the floor/corridor debate in May, in case you are not aware of. https://www.hoover.org/events/currencies-capital-and-central-bank-balances-policy-conference<br /><br />Still trying to dig through the complicated picture, but just a couple of minor comments: in the latest minutes the Fed seems to suggest that the rise in o/n repo rate is associated with a substantial increase in the net supply of Treasury bills. Compare the numbers (SOMA holdings against T-Bill net supply), my guess is the latter indeed might have a larger impact on the amount of safe collateral. <br /><br />In terms of the repatriation story, while it's true that the offshore earnings were pretty much already invested in Treasury securities, these position may unwind as the repatriated earnings can be used to fund, say, merger/acquisition and buybacks. I guess this also alleviates the shortage of collateral a bit. <br />Haonan Zhounoreply@blogger.comtag:blogger.com,1999:blog-2499715909956774229.post-25999716883433802102018-07-05T08:49:20.590-07:002018-07-05T08:49:20.590-07:00Sorry, I was just pointing out what's in the f...Sorry, I was just pointing out what's in the fifth chart. Anything affecting the supply and demand for Treasuries (and close substitutes) will matter. Tax repatriation may make no difference, as much of that is just accounting tricks that move the tax home, with no implications for where the assets actually reside, or the demand for liquid safe assets. In the fifth chart, the increase in reverse repos issued by the Fed and held abroad should reduce the demand for Treasuries, thus working in the right direction, but the timing is wrong. The increase happened in early 2016.Stephen Williamsonhttps://www.blogger.com/profile/01434465858419028592noreply@blogger.comtag:blogger.com,1999:blog-2499715909956774229.post-21410016137072296312018-07-05T07:31:14.201-07:002018-07-05T07:31:14.201-07:00"I think the reverse repos held abroad...&quo..."I think the reverse repos held abroad..."<br /><br />Not reverse repo. Companies are holding Treasuries and other highly rated bonds overseas.<br /><br />"But that wouldn't explain why the overnight rates move up to the top of the band early this year."<br /><br />Are you questioning the timing? Tax reform legislation only came into effect last December, and there would probably be some delay as companies devised strategies for repatriation and the selling off of their Treasury positions. That would be consistent with the GC repo rate moving up to be in-line with fed funds/IOER later in spring rather than earlier.JP Koninghttps://www.blogger.com/profile/02559687323828006535noreply@blogger.comtag:blogger.com,1999:blog-2499715909956774229.post-32631659661235361962018-07-05T04:11:59.738-07:002018-07-05T04:11:59.738-07:00Given the large balance sheet, it would have been ...Given the large balance sheet, it would have been better if the ON-RRP rate had been set equal to IOER. That would have mitigated the problem. Though then QE starts looking silly. Take Treasuries and MBS out of the market so they can't be used in the repo market, then have the Fed use them in the repo market.<br /><br />Don't know if your tax repatriation story works. I think the reverse repos held abroad takes pressure off the Treasury market. But that wouldn't explain why the overnight rates move up to the top of the band early this year.Stephen Williamsonhttps://www.blogger.com/profile/01434465858419028592noreply@blogger.comtag:blogger.com,1999:blog-2499715909956774229.post-91309139439553707382018-07-04T20:38:16.272-07:002018-07-04T20:38:16.272-07:00Great post, Steve.
So is it now fair to say that...Great post, Steve. <br /><br />So is it now fair to say that the whole ON-RRP program was overkill? You say that the ON-RRP program could be discontinued, and it would make no difference. Maybe that was always the case i.e. maybe the fed funds rate always would have gone up when IOER went up, even if the ON-RPP had never been implemented. <br /><br />"Why is this happening? Well, you don't have to think too hard to figure that out. The tightening up of the overnight market coincides with the phasing-out of the Fed's reinvestment program...primary fodder for the overnight repo market. "<br /><br />Ok, here's another theory for you. Maybe Trump's tax repatriation program is freeing up a bunch of U.S. Treasuries that had previously been frozen in overseas jurisdictions for tax reasons. Companies can now sell their hoards to honor tax obligations, reduce debt, reward shareholders, invest in projects, pursue acquisitions, etc. The buyers of will be financial institutions who had previously been starved of Treasury collateral. JP Koninghttps://www.blogger.com/profile/02559687323828006535noreply@blogger.com