Invesaro publishes investment theses and settles them against the market in public: with a date, a basket of companies and a verdict.
Every HBM stack, the heart of AI accelerators, is far less dense than a commodity DRAM die: the same bit capacity takes up nearly three times the silicon area, because you add logic layers, TSV interconnects and worse yields. The upshot is that as Samsung, SK Hynix and Micron redirect more and more wafer capacity to HBM and to the newest DDR5, a disproportionately large chunk of commodity memory supply disappears from the market.
Each one is a statement that has to be true. We go back to every live thesis on a schedule and check these against public sources, so each leg carries either what we found or the date we look next.
Archive: reviews of this thesis are paused. Last review September 26, 2026.
Samsung, SK Hynix and Micron publish no announcement reversing, extending or postponing their DDR3/DDR4 end-of-life and legacy-node shutdown plans before 6 January 2027.
Checked September 26, 2026, nothing published either way
DDR4 contract prices quoted for December 2026 are at or above their August 2026 level, rather than falling back toward the pre-shortage 2025 range.
Checked September 26, 2026, nothing published either way
Nanya Technology reports a consolidated gross margin above 15% in its Q3 2026 quarterly result and above 15% again in its Q4 2026 result.
Checked September 26, 2026, nothing published either way
CXMT brings no new DRAM fab or expansion phase online before 6 January 2027 that adds more than 50,000 wafer starts per month of DDR4/DDR5 capacity.
Checked September 26, 2026, nothing published either way
Winbond reports monthly revenue at least 15% above the year-earlier month in at least three of the months from September 2026 through January 2027.
Checked September 26, 2026, nothing published either way
What has happened since publication. Every entry is computed as of its own date, from price history.
We went back to this thesis on day 48 of 150 and checked each of its 5 assumptions against public sources. No news either way: 5. The equal-weight basket stands at +3.0% since publication day. The benchmark (.INX) returned -2.3% over the same window, so the basket is ahead of its market by 5.3 percentage points. 102 days remain until the verdict.
The first quarter of the horizon is behind this thesis. The equal-weight basket stands at -2.1% since publication day. The benchmark (S&P 500) returned -2.2% over the same window, so the thesis is moving with its market. The strongest contributor is 2408 (+1.9%), the weakest is 2337 (-5.8%). 112 days remain until the verdict.
We went back to this thesis on day 33 of 150 and checked each of its 5 assumptions against public sources. No news either way: 5. The equal-weight basket stands at +3.6% since publication day. The benchmark (.INX) returned -2.1% over the same window, so the basket is ahead of its market by 5.7 percentage points. 117 days remain until the verdict.
We went back to this thesis on day 19 of 150 and checked each of its 5 assumptions against public sources. No news either way: 5. The equal-weight basket stands at +12.4% since publication day. The benchmark (.INX) returned -0.3% over the same window, so the basket is ahead of its market by 12.7 percentage points. 131 days remain until the verdict.
This basket is at +11.0% since publication, 4 days after the thesis was posted. The benchmark (S&P 500) returned +0.5% over the same window, so the basket is ahead of its market by 10.5 percentage points. The strongest contributor is 2408 (+12.5%), the weakest is 2344 (+8.3%). This is not a change of status: the horizon keeps running and the verdict only lands once it expires.
This is analysis, not investment advice and not a recommendation to buy or sell anything. We publish it and track it in public, mistakes included. Any decision is yours and yours alone.