Tightening Financial Conditions

Disorderly Disinversion

As 10-year Treasuries (USTs) decisively broke above the October ‘22 closing high of 4.24% and real 10s (TIPS) surged towards 2%, their highest level since the financial crisis, the explanatory narrative evolved from increased Treasury supply and reduced foreign demand to stronger growth. Our long-standing view is that since the global savings glut, when China and Japan accumulated large UST positions to suppress their exchange rates, and later due to QE, the market implied growth outlook is a tertiary factor in determining real rates. While retail sales decidedly beat consensus forecasts, as previewed last week, the August NAHB Housing Market Index slipped back to the 50 boom/bust line, much weaker than expected, July housing starts were in line, the first two August regional Fed manufacturing surveys were mixed, while July industrial production was strong. The Atlanta Fed GDP Now tracking estimate jumped after retail sales, but we view this indicator as noise at this point (1 month of data) of each quarterly cycle. The control portion of retail sales (ex-autos, gas & building materials) is roughly 20% of personal consumption expenditures (PCE), we have no data on services (2/3’s of PCE), nothing on trade or capital investment. We do think the gross domestic income/earnings recession is ending, but the macro and micro (retailer earnings) data this week did little to change our view. Another way to illustrate this point is with the NY Fed’s weekly index (WEI), here is their commentary explaining the weekly decline.

“The decrease in the WEI for the week of August 12 (relative to the final estimate for the week of August 5) is due to falls in steel production, tax withholding, consumer confidence, railroad traffic, and fuel sales, which more than offset increases in retail sales and electricity output and a decrease in initial unemployment insurance claims.”

What was significant, and likely played a larger role in the continuation of the real rate bear yield curve steepening, was yen and yuan exchange rate weakness. Following a round of decidedly negative Chinese industrial production, fixed asset investment, retail sales and new house prices, as well as struggles with a major real estate developer and nonbank credit provider (trust manager), the PBoC intervened to slow the decline in the yuan. The stale June Treasury International Capital Flow report showed China continuing to reduce their holdings, and selling dollars to support the yuan is a clear signal they have more USTs for sale. Given that Treasury Secretary Yellen’s best customers — the Fed, banks and export dependent Asian sovereigns —are better sellers, she needs lower prices to convince hedge funds to cover/unwind steepeners and unlevered pension funds, money managers and US households to extend duration to absorb the extra $500 billion of coupon securities she has for sale after Labor Day.

There are two major events between now and Labor Day that could alter the outlook and potentially halt the unfolding risk-off episode. First up is the Kansas City Fed’s 2023 Economic Policy Symposium, "Structural Shifts in the Global Economy," Aug. 24-26. The second is an early employment week this month, with July JOLTS and the August Conference Board Labor Differential on Tuesday August 29 and the August Employment Situation report on Friday September 1. The first look at 2Q gross domestic income and the August personal consumption deflator report on August 30 and 31 could solidify the disinflation and improving growth narrative but are unlikely to significantly alter the outlook. On balance the Treasury market move is looking extended, but equities are lagging, implying until and unless the S&P gets closer to 4300, the velocity of the decline accelerates, and our measures of cross asset class risk increase, we would keep our powder dry. Later in the note we will discuss a macro doom loop that we would consider expressing through long energy and short banks.

Figure 1: The upside breakout of real rates is an issue for Treasury Secretary Yellen as she prepares to finance the most profligate Administration just FDR’s wartime government. This reasonably high velocity move is overcoming rate suppression from the Fed’s still massive holdings.

Rate Sensitivity

Stronger than expected July retail sales and decent results from large retailers excited consumer-obsessed financial media hosts and their guests, but we expected decent results. This was in part due to a persistent rise in consumer discretionary net revisions, as well as, in our view, that the household and large nonfinancial corporate sectors have termed out their debt and improved their cash flow, leaving them largely immunized from tighter monetary policy. The debate about excess savings is not meaningless, but it is only a minor factor in the broader consumer outlook, in our view. Improving cash flow due to disinflation and improving productivity leading to positive real income growth is a positive factor. Another plus is 15 years of household sector deleveraging (notwithstanding the clickbait headlines about record aggregate levels of debt after last week’s NY Federal Reserve quarterly report) and interest obligations remain the lowest levels since the Fed began tracking interest as a percent of disposable income in 1980. While the below chart, constructed by Macrobond, our macro database provider, shows the rate of change of fiscal support easing, the aggregate level of fiscal transfers to individuals relative to disposable income is 100bp higher than it was pre-pandemic, and CBO projects transfer payments are a major source of record levels of government spending to GDP over the 10-year forecast horizon.

Figure 2: Thanks to Macrobond for this chart, one important note, before fiscal stimulus created excess savings, trend consumption was 3%.

Strong performance for nonstore retailers (ecommerce) may be a sign that goods for services demand is stabilizing. Our trend analysis shows excess services demand of $252 billion, 1.7% of personal consumption expenditures, and goods consumption $77 billion below trend. Our measure of real retail sales (mostly goods, though dining out was strong) shows the 3-month annualized rate marginally above the long-term trendline, hinting that goods demand is improving. Another trend evident in the retail sales series is a post-pandemic level shift and rate of change acceleration for ecommerce market share. Nonstore retailer sales share of total sales was 14.6% in February ‘20, it spiked at 20.7% in April ‘20 during the depths of the lockdowns, pulled back to 16.8% in July ‘21, before accelerating to 19.4% in July ‘23. Our bias is to conclude that this trend is a net positive for labor productivity.

The risk to the rosy outlook for consumer spending is our unstable equilibrium thesis: that is the deeply inverted yield curve tightens bank credit for the small business sector, which leads to lower employment. In other words, long and variable lags. For now, rising revisions and margins for the consumer discretionary sector are compelling evidence the household sector remains immune to tighter monetary policy.

Figure 3: Real retail sales are marginally above the longer-term average, perhaps due to a recovery in goods demand.

One way to illustrate interest rate sensitivity is through the duration of fixed income sectors. Duration of the mortgage-backed securities index is at record highs, given the post-GFC market share gains for Fannie, Freddie (GSEs) and the original subprime originator, FHA, the index duration highlights the household sector’s low sensitivity to higher mortgage rates. The dichotomy between the duration of the corporate aggregate index of 7 years, off the pandemic highs but considerably longer than the pre-financial crisis trend either side of 6 years, and the high yield index at index lows of 3 1/2 years, is consistent with our view that large corporates termed out their debt but businesses without access to long term investment grade credit are vulnerable both in terms of price as well as the supply of credit. The extended duration of the debt of these two large sectors argues for longer policy lags, rather than shorter lags Governor Waller asserted was a consequence of Fed policy communication.

We have long argued that policy support for 30-year fixed mortgages, a product that couldn’t exist without the presence of the GSEs, impairs monetary policy transmission. Following the financial crisis, negative equity made a large portion of the mortgage universe unrefinanceable at a time when household debt and financial obligations ratios were at all-time highs. Today, when the Fed attempted to reduce demand to ease price pressures, fixed rate mortgages made the household sector largely immune from policy rate. The insensitivity to hikes was exacerbated by the Fed’s approach to hiking, namely passive balance sheet contraction. Recall the three channels of large-scale asset purchases: liquidity (bank reserves), duration (yield curve flattening) and volatility suppression (mortgage prepayment risk). Passive QT has only now reached the point where liquidity is tightening, the 3m10y Treasury curve inversion was only deeper twice in US history (‘79 and ‘80) and implied volatility remains suppressed due to the vast majority of the mortgage universe unrefinanceable (out-of-the-money). One way to illustrate this point is 65% of the Fed’s agency mortgage-backed securities holdings are 2% and 2.5% coupons, the yield on the Fannie Mae 30-year fixed rate current coupon security is 6.12%.

Figure 4: Duration of the mortgage-backed securities index is at an all-time high, the corporate agg index was steadily increasing pre-pandemic, it spiked and is back to the pre-pandemic 25 year high. High yield is shortening. The dichotomy illustrates our view that the supply of credit is contracting.

Much has been written about office building delinquencies and rollover risk in ‘24 and ‘25. In an interview right after ours on CNBC this week, the CEO of RXR Realty, Scott Rechler, made a very cogent argument that the larger risk for small and medium sized banks is in the apartment sector. Others, including Mike Green (on our Substack recommended list), have written about a coming glut of multifamily rentals. This shouldn’t surprise anyone given the increasing role of the GSEs in this space. Like office buildings, a deeply inverted curve reduces the supply of credit and if the curve disinverts the bear way, higher long rates, when construction loans turn into permeant finance with slower rental rate increases, many deals are not going to work.

Figure 5: Bank, apartment and office REIT credit spreads widened sharply in March following the collapse of SVB, but interestingly began widening when the Fed started the rate hike cycle.

The Petrodollar Doom Loop

Our strategic financial sector weighting is neutral, however tactically we are leaning short. The ‘real’ (rate) bear steepening is problematic for the 30% of bank assets invested in securities. Those holdings will likely be a focus of 3Q earnings absent a major Treasury and mortgage-backed securities rally by the end of September. The Fed is not expected to hike in September, however rising deposits at small banks relative to larger institutions imply rising costs of goods sold (deposit betas/rates). Flows into money market funds averaged $27.5 billion over the last four weeks, the same pace since the collapse of Silicon Valley Bank, way above the $1.8 billion pre-SVB early 2023 pace. These flows imply the pressure on deposits persists, in the latest weekly Fed H.8 report large bank deposits contracted $28.7 billion and small bank deposits increased $5.7 billion, we suspect the price of small bank financing is rich. Bank credit is contracting (-0.90%), the composition highlights our unstable equilibrium thesis. Securities holdings are 11.2% below a year ago, loan growth is +4.5%, consumer loans +5.8%, residential real estate +5.3% but commercial & industrial (C&I) loan growth is -0.3% from a year ago. Small bank C&I just crossed under deposits in the latest week and is contracting 1.6% annualized. On an investor call this week we were asked about the magnitude of Fed rate cuts to stabilize the banking system. While it appears the r** (financial stability neutral rate) for bank securities portfolios appears to be just above 3%, the $1.7 trillion of USTs purchased by the banking system in ‘20 and ‘21 likely had a duration similar to the Fed’s 6-year average maturity. Consequently, ~60% of those holdings are now 3-year USTs and the other 40% are 2s. The nearly $1 trillion of mortgage-backed securities are going to hang around for quite a bit longer but the combination implies a rising r**. We suspect a policy rate in the low 4% range would be sufficient for the banking system to earn their way out of their underwater securities holdings.

Figure 6: It looks like the blue wire (credit) and the black wire (deposits) might cross; it could be explosive.

Meanwhile, across the world’s biggest pond, Asian goods exports and energy import dependent state directed economic systems are struggling with global supply chain restructuring and an oil market that is increasingly tight. Our former colleague Helima Croft, in a CNBC interview Friday morning, suggested Saudi supply cuts, combined with improving US and surprisingly Chinese demand, points to higher prices. The dollar’s evolution into a petrocurrency is exacerbating the rising energy costs, thereby forcing the PBoC to sell dollars held in USTs to stabilize their exchange rate, with others like the BOJ likely to follow. The BOJ has an alternative to exchange rate intervention, they could loosen yield curve control further, but both paths led to the same destination, additional UST sales. This doom loop puts additional pressure on US banks. On the flip side, energy stocks just lapped the earnings comps that were flattered by the Russian Invasion price spike, and valuation is low despite shale lowering the elasticity of supply leading to higher and more stable return on invested capital. Long energy and short banks should capitalize on the bear steepening doom loop at least until the Fed capitulates and there is a clear path to a disinflation driven bull steepening.

Figure 7: This data does not reflect developments in August. We suspect the Euro Area buying is largely attributable to energy exporters working through European banks.

Structural Shifts in the Global Economy

One of the structural shifts in the US economy in dispute, as evidenced by the considerable dispersion of ‘25 policy DOTS from 5.75% to 2.5%, is whether r* is higher post-pandemic that the globalization, demographics and private sector deleveraging driven low r* of the last two decades. While this sounds mostly theoretical, a key member of the leadership, NY Fed President Williams, is on the record with a view that r* remains low implying policy will become excessively restrictive due to disinflation in ‘24. If this view spreads within the FOMC, the outlook for ‘24 rate cuts will improve. We suspect there may be a decent debate about the impact of deglobalization on productivity. Former Vice Chair Brainard gave a speech when she was still on the board concluding deglobalization would reduce productivity, and we disagree. China’s ascent to the world’s factory was largely due to substitution of labor for capital, OECD productivity and unit labor cost measures imply highly automated capital for labor production will improve, not degrade productivity growth. The most important question is what Chairman Powell’s core message will be in his Friday morning speech. No doubt they want a third consecutive 0.2% core CPI monthly increase before ruling out a September hike. Consequently, the Chairman is likely to retain optionality. We think he should reframe inflation dynamics from his core goods, rent of shelter and sticky core services less rent of shelter due to unsustainably high wage growth framework speech last November. At a minimum he could hedge by speculating cooler labor demand and resilient economic output imply faster productivity growth, thereby making 4% wage growth sustainable. Chairman Powell’s first Jackson Hole speech in 2018 was about humility given movement in the celestial stars (r* and u*, equilibrium interest and unemployment rates). It is time for some additional humility, the political pressure is about to intensify.

“Participants remarked that the labor market continued to be very tight but pointed to signs that demand and supply were coming into better balance. They noted evidence that labor demand was easing—including declines in job openings, lower quits rates, more part-time work, slower growth in hours worked, higher unemployment insurance claims, and more moderate rates of nominal wage growth. In addition, they remarked on indications of increasing labor supply, including a further rise in the prime-age participation rate to a post-pandemic high. Participants also observed, however, that although growth in payrolls had slowed recently, it continued to exceed values consistent over time with an unchanged unemployment rate, and that nominal wages were still rising at rates above levels assessed to be consistent with the sustained achievement of the Committee’s 2 percent inflation objective. Participants judged that further progress toward a balancing of demand and supply in the labor market was needed, and they expected that additional softening in labor market conditions would take place over time.”

Figure 8: Trends in revisions are a leading indicator, we’ve been arguing all year that demand for labor is considerably weaker than the headline payroll gains imply.

Tactics: Over? Did you say the risk-off episode is over?

As the real bear steepening began, we penned a decisive break above 4.25% target for 10-year USTs, 2% for real 10s and a 5-7% S&P 500 pullback. While we are in the range of all of those targets, we expect Chairman Powell to retain optionality next Friday, we are two weeks away from the next employment report and 3 1/2 weeks from the pivotal August CPI report. Finally, our cross-asset measures of risk are at their long-run median levels implying the investor positioning is some way from levels that would trigger a reversal. The VIX is spot on its long run median, the term structure is normal, skew is elevated, and implied correlation is low. The MOVE Index is 1 standard deviation elevated, this is consistent with USTs leading equities lower, and approaching our view of current fair value (2% 10-year TIPS). To be clear, we still view 10s as structurally overvalued due to Fed holdings as evidenced by a -56bp term premium. Despite pressure on the yen and yuan, the JP Morgan FX Volatility Index is below its longer run median. Friday morning’s bounce in USTs as equities declined is a decent sign, but we suspect both markets will be back under pressure next week ahead of Jackson Hole.

Figure 9: Our measures of risk are one of our critical positioning/sentiment indicators, it’s what they pay (for protection), not what they say (surveys).

Figure 10: There is a change, we are describing our view on financials as underweight. We need visibility on a reversal of the last three excessive rate hikes for the pressure to ease on the sector.

Key Investable Themes & Asset Allocation:

• Deglobalization & Capital Spending Boom: Industrials XLI, SMH

• Technology Innovation Diffusion: Healthcare IYH, Industrials XLI

• Credit storm tactical trade: Long RSP (equal weighted S&P), short SPY

• Global Equity Allocation: Overweight US and UK equities SPY, EWU, underweight export dependent economies (China FXI, Germany EWG)

• US Asset Allocation: Overweight equities SPY, underweight Treasuries TLT, overweight credit LQD

• Getting closer to another tactical long in USTs. We like the LQD Aug 106-102 put spread as a way to play spread widening on investment grade credit resulting from heavy Treasury issuance.

Barry C. Knapp

Managing Partner

Director of Research

Ironsides Macroeconomics LLC

908-821-7584

https://ironsidesmacro.substack.com

https://www.linkedin.com/in/barry-c-knapp/

@barryknapp