13-Dec-2023 FOMC meeting
17-Dec-2023. Confused by the FED? December 1: J. Powell: “It would be premature to ... speculate on when policy might ease”, December 13: J. Powell: Rate cuts are something that “begins to come into view” and “clearly is a topic of discussion”, December 15: J. Williams: “I just think it's just premature to be even thinking about (rate cuts)”. So it's rather difficult to rely on what the Fed says... It is also difficult to rely on their forecasts regarding the future policy path. Figure 1 shows the latest median path with the range of the average forecast error plotted (average error based on 48 policy paths published to date since 2012). Link to the analysis below the text. It is also difficult to rely on basic economic data such as the change in nominal GDP or the change in core inflation (Figure 2). In this case, the Fed appears to be responding quickly to lower GDP growth (2008, 2020), but is not willing to respond to higher GDP and higher inflation (unless it no longer has a choice...). It is also difficult to rely on the Taylor's Rule (Figure 3), which in Q4 2021 suggested the FED's rate at 8.5% - while the FED kept insisting that inflation was temporary, keeping rates at zero and conducting QE. Go figure. Trade accordingly.
14-Dec-2023. FED’s Policy in 2024.. There ya have it folks... After yesterday's FOMC meeting, markets received a large dose of clarity regarding the Fed's policy in 2024. Just look at yesterday's reactions of the stock market (S&P500 +1.37%), US rates (UST 10y: -18 bps; 2y: -30bps) and the dollar index (-0.95%). As far as FED's dot-plot is concerned.. the peak hawkishness was on September 20, 2023. Figure 1. The latest path from December 13 is much lower... this is one of the reasons for such a market reaction... That's all fine, but what are the chances that inflation will spoil the party in the second half of 2024 / 2025? So what about fiscal policy in 2024? Unfortunately, we don't have a dot-plot here... but the latest data for November shows spending increased as a % of nominal GDP. In November, Total Federal Outlays amounted to $588.8 billion (+17.6% YoY), while the sum of these expenditures for the last 12 months amounted to $6.29 trillion and represented 22.7% of today's nominal GDP. Figure 2 shows this series against the annual change in inflation. The current 22.7% looks safe... but in March 2021 it was 33.7% of GDP! What is the pre-pandemic norm? The average for 2014-2020 is 20%. What’s the average annual change in inflation for 2014-2020... try… 1.56%. See Figure 3. Figure 3 shows government spending moved forward by 13 months... against the annual change in inflation. The question is, what are the chances of returning to 20% next year or any time soon?
Taylor rule
20-Dec-2023. Taylor Rule. Part 1. The Taylor Rule/Principle is a popular approach to estimating the central bank interest rate. It was proposed by Professor John Taylor of Stanford University in 1993 (link below), and can come in many varieties, but generally we can think of it as the sum of three variables: (i) Long-term central bank interest rate, (ii) +/- Correction due to current inflation above/below the inflation target, (iii) +/- Correction due to current GDP (growth) above/below potential GDP (output gap). Today a little bit about point (i). Nowadays it is widely accepted that the ideal (target) level of inflation is 2%. But the central bank rate cannot be at 2%, because this would mean that the real central bank rate would be 0% (doesn’t make sense). Therefore, to inflation target rate of 2% we should add the natural real interest rate, i.e. r*. What is r*? If we look at the latest FOMC economic projections (link below), the median expected longer run inflation is 2.0%, and the federal funds rate longer run is 2.5%... hence the conclusion that r* is (according to such a projection) 0.5%. This is just one method of estimating r*. FOMC projections are only available from 2012, so to calculate r* over a longer period of time, we can use various models. But where should the Fed rate be if we assume that (i) inflation is on target and (ii) economic growth is equal to potential GDP growth. Figure 1 shows this rate over time. Interesting conclusions can be drawn, e.g. that since 2010 we have had (as a combined effect) inflation below the target and a negative output gap (growth below potential). Of course, if inflation is above the target, the FED rate should increase, and similarly if GDP growth is above potential, the FED rate should additionally be higher. Figure 2 is a similar chart but from 1954. In this case, R* is the average of the estimates of 3 different models (see the chart for details). Figure 3 shows different estimates of the real natural interest rate r* over time. Below a link to a previous publication about Taylor Rule.
FOMC projection error
14-Dec-2023. Policy path projections of Federal Reserve Board members. Since 2012, we have already had 48 projections of the future policy path available. What is their accuracy vs actual future interest rate? Figure 1 shows the average forecast error depending on the number of quarters that have passed since the FOMC meeting. After 1 year, the average error is 1.1 percentage points (maximum error 4.3 points), After 2 years the average error is 1.7 points (max 5.3) After 3 years the average error is 1.7 points (max 3.0). In the case of the "Longer run" forecast, I assumed that each such forecast is 2 years ahead of the longest available forecast at the end of the calendar year. Figure 2 shows the most accurate projection in data history (FOMC as of June 15, 2016). Figure 3 shows the least accurate projections (FOMC of September 26, 2018 and September 16, 2020).
different fed funds paths
7-Nov-2023. Different Fed Funds Rate paths. The chart shows various paths of the future Federal Funds Rate: 1) from the moment "something just broke" - which took place on March 24, 2023, at the epicenter of the crisis of US regional banks, 2) until "nothing will break in the near future” or max higher-for-longer narrative - i.e. the most aggressive valuation of the FED’s rate path by the market (October 17, 2023). But the FOMC itself was even more aggressive with its projection as of September 20, 2023. Either way, the market has always had trouble predicting the future path of the Fed Funds Rate.
how long can the pause last?
9-Nov-2023. How long can the Fed's pause last? In 2018-19, the pause lasted 7 months, then there were 3 rate cuts and a recession (6.5 months after the end of the pause). In 2006-07, the pause lasted 14.5 months, then there were 2 rate cuts (by a total of 75 bps) and a recession (3 months after the end of the pause). In 2000-01, the pause lasted 7.5 months, then there were 2 rate cuts (by a total of 100 bps) and a recession (2.5 months after the end of the pause). In 1989, the pause lasted 3.5 months, then there were 5 rate cuts (by a total of 150 bps) and a recession (13 months after the end of the pause). According to Fed Funds futures contracts, the first rate cut will be June 12, 2024 - then the pause would last 10.5 months. Details on the chart.
FOMC: 20-Sep-2023
20-Sep-2023. Today's FOMC: rates unchanged as expected, but the median projected rate in 2024 is "unfortunately" as much as 50 bps higher than in the June projection: it is really H4L (higher for longer). Today's market reaction (from the announcement of the FOMC decision to the close): 10Y: +8 bps (hawkish) 2Y: +10 bps (hawkish) S&P500: -1.16% (hawkish) EURUSD: -0.60% (hawkish) But on the other hand, the FOMC almost never gets it right with its projections (see the chart with projections since 2012 against the actual Fed Funds Rate) - so FOMC projections shouldn't matter too much beyond the short-term reaction of markets.
post-FOMC fallout: Day 2
22-Sep-2023. Post-FOMC fallout. Day 2. Summary of the main asset classes. Day 1 (from the announcement of the FOMC decision to the close) + Day 2 (at close) + Day 3 (intraday, this morning) = total US Treasuries: massive bear-steepening: 10Y: +8bps +10bps + 0bps = +18bps 2Y: +10bps - 3bps + 0bps = +7bps US Equities, hawkish: S&P500: -1.15% -1.64% +0.2% = -2.6% Nasdaq100: -1.43% -1.84% +0.3% = -3.0% EURUSD: -0.59% -0.04% -0.07% = -0.70% (hawkish) Two main conclusions: 1) The market reaction is like in a classic inflationary regime as in the cycles of the 1970s: stocks and bonds fall together, see the two charts for the S&P500 and 10y UST respectively. Currently, according to the market, we already have a pause (the probability of an increase on November 1 is only 26%, and on December 13… 45%), and the first rate cut will only take place at the FOMC meeting on July 31, 2024. 2) It is better for the American consumer to slow down relatively quickly, otherwise the FED is definitely going to “out-hawk” RoW (rest of the word), and the strengthening USD will only contribute to more troubles ahead for the RoW in the first instance (see the chart of the main currency pairs).
