Growth Expectations Are Probably Too High | Andy Constan on Post-Midterms Macro View, Bond Supply Issue, and QRA (Quarterly Refunding Announcement)

The midterm result makes major fiscal expansion unlikely, Andy Constan said; absent a crisis, political gridlock should leave fiscal policy broadly unchanged. He sees current fiscal policy as a slight drag on growth because government spending net of taxes is not expected to grow faster than GDP, leaving the private sector to carry the economy. A major slowdown or financial crisis could produce bipartisan deficit spending, but he does not expect either. He also sees no reliable modern-era pattern of market gains around midterms; elections matter more when they bring substantive policy changes.

Constan expects growth expectations to weaken further over the next three to six months after being marked down in the preceding three to six weeks, a change that helped drive the bond rally. He expects disappointment from expectations he called “pretty damn high,” but not necessarily a recession. Strong recent consumption and AI investment have supported growth, though consumption remains uneven and has relied more on wealth, investment and interest income, and low-rate borrowing than wage gains. Imported AI capital equipment limits its direct contribution to U.S. GDP. Without economic weakness, he sees little basis for 10-year yields to fall to 4.5% soon; strong growth, above-target inflation and the large deficit all constrain an easy decline.

For bonds, Constan treats growth and inflation expectations as the primary yield drivers, with supply and demand offering an additional trading edge. Strong growth can lift real yields by making investment more attractive than lending. He sees government and corporate borrowing, AI-related issuance, and prospective European and Japanese defense spending adding to bond supply and risk premiums. In his view, borrowing has shifted conditions from a long-running savings glut toward a savings deficit. Supply does not ensure failed auctions, since prices can adjust to attract buyers, but the financing required to sustain current growth expectations may not clear at current prices.

The QRA left planned coupon issuance unchanged and continued Treasury buybacks of 20- to 30-year bonds at an annual pace of $192 billion, up from about $64 billion. Constan estimated net annual supply of those longer-term bonds at roughly $252 billion, substantially below prior years, and called the announcement broadly expected and a market non-event. He said the Treasury has issued about $4.2 trillion a year in coupon auctions, TIPS and floating-rate notes, while roughly $7.5 trillion in bills rolls through auctions, for total annual debt sales of about $12 trillion to $13 trillion against a deficit near $2 trillion. Buybacks suppress long-term rates, but the QRA left yields free to respond to macro conditions and private-sector supply and demand.

Constan said Treasury buybacks had influenced the long end despite earlier criticism from Treasury Secretary Besson of Janet Yellen’s use of them. The buyback pace for 20- and 30-year bonds had risen to $192 billion annualized as of August 18, and he said the effort to suppress rates had not been intensified in the latest QRA. He does not think it has been abandoned: if 10-year yields reach 6% and stocks fall, he expects the Treasury to reduce supply or increase buybacks. After building a maximum long-bond position in early October, he was flattening it and awaiting the election outcome and QRA; he still likes bonds as portfolio holdings, but sees less appeal in a speculative trade after yields fell from their highs.

Constan has been neutral on stocks for months as major indexes moved sideways and low realized volatility made options trading difficult. He said S&P 500 EPS rose about 33% this year, with roughly 16% growth expected next year, while earnings expectations remain extraordinarily high; EPS can also be boosted by nonrecurring valuation gains, such as Amazon’s gains from investing in SpaceX. Buybacks have reduced outstanding shares by about 2% of GDP for decades, but projected net share-count reduction falls to zero next year, a shift he estimates at roughly $700 billion in net stock supply and demand. He sees no aggregate Nasdaq bubble pop, but thinks the market may struggle to meet its built-in growth expectations.

He does not consider the AI trade finished or AI tokens and semiconductors worthless, describing AI as a major productivity opportunity that can still face setbacks; he acknowledged that his recent bullish semiconductor view was wrong. The shift away from semiconductors and neoclouds toward companies that overspent on capital expenditure reflects expectations of less compute capacity. Scarcity could raise prices, but it is unclear whether higher prices would offset lower sales volumes or whether providers can pass costs through to AI-token customers. If demand slows, infrastructure and chip sellers could face lower sales and a glut. Separately, he noted that buybacks have been cut by major AI investors and public-market issuance has become a remaining funding source for investment.