Hi! I reran the model with the two corrections from the thread. CSM rebates are no longer added to DAO profit on top of reported revenue, and I now show DAO profit in two versions, with treasury income as in the report and with treasury income set to zero. Since that line can turn negative, the zero case is the more cautious one, and if it is actually negative the profit figures would be lower still. Buybacks do not depend on this assumption, because NEST does not see treasury income at all, so only the profit and risk columns change between the two versions. Below are the two calculations discussed above, a budget table and a yearly reset of the NEST counter.
1. Buybacks at different budgets, with the threshold equal to the budget
Share stays at 50%, the starting deficit is the actual −$542,451, and the horizon is 12 months from 1 October 2026 across 1,000 ETH price paths. Actual expenses are assumed to match the approved budget.
The average is taken over all 1,000 scenarios, including those with no purchases at all, so it should not be read as what NEST will buy in a normal year. It is pulled up by the scenarios where ETH rises and NEST buys a lot. That is why I also show the median, the outcome of a typical year, and the chance of making any purchase. At the current budget the median is zero, meaning that in most scenarios NEST buys nothing over the next 12 months, and the $1.74M average comes entirely from the stronger ETH paths. Around $35M the median turns clearly positive, and at $30M a typical year already brings about $3.2M of buybacks.
| Budget = threshold | Average buybacks | Buybacks in a typical year (median) | Chance of any purchase | DAO profit before buybacks, with / without treasury income | Share of purchases in loss-making years, with / without treasury income |
|---|---|---|---|---|---|
| $43.8M (as now) | $1.74M | $0 | 36% | $1.0M / −$1.8M | 7% / 10% |
| $40M | $2.32M | $0 | 47% | $4.5M / $1.7M | 4% / 12% |
| $38M | $2.69M | $0.39M | 56% | $6.3M / $3.5M | 7% / 13% |
| $35M | $3.35M | $1.26M | 69% | $9.0M / $6.2M | 7% / 15% |
| $32M | $4.13M | $2.35M | 82% | $11.8M / $8.9M | 7% / 14% |
| $30M | $4.71M | $3.18M | 89% | $13.6M / $10.8M | 6% / 13% |
| Current NEST ($40M threshold, $43.8M spending) | $2.32M | $0 | 47% | $1.0M / −$1.8M | 14% / 26% |
Every $2M taken out of the budget adds roughly $0.3–0.6M of buybacks a year, and the effect grows as the budget gets smaller. The $40M row buys exactly as much as NEST does today, but with a much smaller share of purchases in loss-making years, because the threshold then matches what the DAO actually spends. Without treasury income the picture gets stricter, but the conclusion stays the same. At the current budget the DAO is slightly in the red before any buybacks, and today’s NEST makes about a quarter of its purchases in loss-making years. With the threshold equal to the budget that share is roughly half as large at any level of spending, and the cut to $30M still leaves about $10.8M of profit.
I think a table like this belongs in the 2027 budget request itself, so delegates can see what each level of spending means for buybacks at the moment they vote on it.
2. Yearly reset of the counter
Here I kept the current $40M threshold and 50% share and compared three rules over three years. The average again includes years with no purchases.
| Deficit rule | Average buybacks per year | Years with any purchase | Share of purchases in loss-making years, with / without treasury income |
|---|---|---|---|
| Accumulates (current) | $2.74M | 42% | 14% / 20% |
| Reset every year | $2.96M | 51% | 21% / 31% |
| Reset every quarter | $3.05M | 57% | 29% / 38% |
A yearly reset adds about $0.2M of buybacks a year, roughly 8%, and makes NEST buy in more years. The price is that a purchase is about one and a half times as likely to land in a year the DAO ends at a loss. In dollar terms the amount bought in such years stays small, about 2–6% of the total in every variant, because those purchases are usually modest. A reset on 1 January 2027 would also simply write off the current deficit.
So a reset does not change the picture much in either direction. It makes buybacks a bit more regular but does not make them larger in any meaningful way, and it weakens the protection the deficit gives. The budget lever moves buybacks by millions, the reset by a couple of hundred thousand. If the DAO wants a reset anyway, it makes the most sense once the threshold already follows the budget, so that a fresh start each year coincides with a freshly approved budget.
These figures come from a new set of price paths starting on 1 October, so they differ slightly from the tables in the report. The script and results are in the repo. lido-nest-analysis/nest_research/followup_budget_reset.py at main · 0xkzmch/lido-nest-analysis · GitHub
Happy to run other budget levels or a different reset date if that helps with the request.