Hold and Harvest: MSTR

Live strategies

Objective

To earn a yield on a MicroStrategy position that would otherwise sit idle - part of it paid in more MicroStrategy, the rest in dollars that pay down the loan held against it.

The premise is simple. If you intend to hold MSTR for years, its volatility is not merely something to endure. It is an asset in itself. Price moves several percent most days regardless of direction, and a passive holder captures none of that.

The project borrows a small amount against shares already owned and uses that capital to trade the swings. It now runs as two sleeves. One trades MSTR itself and converts every profit back into shares rather than cash, so the share count grows. The other runs the same ladder across a diversified book of equities, indices and commodities and takes its profit in dollars. The original holding is never sold in either case.

Those dollars have a job. They repay the loan secured against the collateral, so the position de-levers as the strategies run. That is what the diversification is for: the other assets are not there to be accumulated, they are there to generate cash that reduces the debt on the one asset this project does want to hold.

In kind and in dollars were both available for the diversified sleeves, as was using the proceeds to buy more MSTR. Dollars were chosen deliberately, and not because they earn the most. Retaining shares would build a long-term position in a dozen names held for their volatility rather than out of conviction; buying more MSTR would concentrate further into the asset the collateral already is. With borrowing already low against collateral, the marginal value of either is growth - and what this sleeve is for is protection.

This is deliberately a low-leverage design. The borrowing is a fraction of what the collateral could support, because the objective is accumulation over years, not amplified returns over months.

How it works

Four mechanisms, running continuously and automatically.

A trend filter suspends buying during sustained declines rather than averaging into them, and each ladder refuses to open a position whose target would sit above a ceiling set for that asset.

Execution is fully automated and runs continuously. Specific parameters are not published.

Projections

Roughly 15–25% more shares from the MSTR sleeve over the next market cycle, with a central estimate near 19% - and roughly 15–25% a year on the capital deployed in the stock sleeves.

Borrowing sits at around 12% of collateral value, a fraction of the available limit, and falls as the strategies run. The share figure rests on four and a half years of price history. The stock figure rests on a few months, and should be read as an expectation rather than a finding.

What the estimates rest on

Four things, none of them certain.

Cycle timing. Bitcoin has historically moved from cycle bottom to cycle top in roughly 35 months, and MicroStrategy's own last cycle ran 23 months from its December 2022 low to its November 2024 peak. The projection assumes a comparable span, preceded by a few months of decline before a bottom forms. Whether that rhythm still holds is openly disputed. Institutional flows are widely argued to have displaced the halving as the dominant driver.

Observed behaviour. The estimate is built from how the strategy actually performed across four and a half years of real price history, in rising, falling and consolidating markets, rather than from assumptions about how it ought to behave.

Volatility compression. Each bitcoin cycle has been less volatile than the last. The estimate discounts the previous cycle's activity accordingly, which is the largest single downward adjustment applied.

Cycle rates on the stock sleeves. The yield figure is the rate at which those ladders have been completing cycles, net of the spread each one pays, applied to the capital they tie up. It comes from a few months of live trading across a dozen names rather than a full cycle, and it assumes those names keep moving roughly as much as they have. A quiet year produces materially less, and a single ladder can sit idle for months if its trend filter stays red.

What moves the number

How uncertain this is

Very. Running the same assumptions through different plausible price paths produces outcomes from roughly 14% to 34%, a spread of more than two to one, from identical inputs. The range is the honest answer; the central figure is a convenience.

Beyond that sits a larger uncertainty. Every figure here assumes bitcoin recovers and carries MicroStrategy with it. That is an assumption, not a finding. Published 2027 bitcoin targets currently span roughly $58,000 to $1 million, which is less a forecast than an admission that nobody knows.

A delayed recovery is not a problem for the strategy. Shares continue to accrue throughout, and the longer the market spends at lower levels, the more of them accumulate. Cheaper prices mean each completed trade buys more stock, and a longer wait means more trades.

The most productive stretch in the entire record was not the bull run. It was the consolidation around the 2022–23 low, which accrued faster than any part of the rally that followed. A market that stays low and moves sideways is close to the ideal condition for this.

Borrowing costs are already deducted throughout. Interest runs at roughly 6.5–8.5% annually and comes directly out of whatever is harvested.

How this compares to leverage

The obvious alternative for a holder who wants more exposure is to borrow and buy. That works, and in a rising market it works well, but it requires choosing a moment, and it concentrates the entire risk into that one decision. Borrow at the wrong point and the position is liquidated before the thesis has a chance to play out.

This takes the opposite approach: the same borrowed capital, deployed gradually across hundreds of trades rather than in a single position, at a loan-to-value a fraction of what most leveraged buyers run. It needs no view on when the bottom is, and it cannot be wiped out by getting one moment wrong.

The one participant it does not beat outright is someone who borrows and buys precisely at the low. But even there the gap is narrower than it looks, and it closes.

Their advantage is capped. Borrowing a fixed sum at the bottom buys a fixed number of shares. As the price rises their debt shrinks in share terms, but the shares themselves never increase. The position asymptotes and stops. Accrual has no such ceiling; it compounds for as long as the strategy runs.

Given enough time and enough volatility, the strategy overtakes even the leveraged bottom-buyer, in a little over a year at the accrual rate seen around the 2022–23 low, longer in quieter conditions. That crossover is structural rather than a matter of timing: one side is capped, the other compounds. And it arrives having carried a fraction of the leverage along the way.

Risks & limitations

The honest version.

• This is not market neutral. The collateral is MicroStrategy and every strategy position is long. When MSTR falls 80% the collateral falls with it, and a broad selloff takes the diversified book down at the same time. These strategies change the share count and the debt, not the direction.
• Diversification here is partial. The strategy book now spans a dozen assets rather than one, and its profits return as cash that reduces the debt - both real improvements on a single-asset design. But the collateral is still MicroStrategy alone, and equities tend to fall together, so the correlation that matters most is the one that reappears exactly when it is least welcome.
• The cost of a bad decline is unwinding, not liquidation. Borrowing sits at a fraction of the limit and the strategy sleeves hold liquid positions that can be closed to raise cash, so a forced sale of collateral is a remote scenario rather than the operative one. What a severe decline does instead is force those positions shut at a loss, mid-ladder, which is the opposite of what they exist to do. The loan remains secured, and a fall beyond anything in the asset's recorded history would still demand action.
• Tokenised equity carries additional risk. The instrument depends on an issuer holding the underlying shares, a custodian, and smart contracts. Any can fail independently of MicroStrategy itself. The token can be frozen by its issuer.
• The evidence base is thin. The accumulation sleeve rests on one asset across one market cycle, roughly four and a half years, with parameters chosen in knowledge of that history - which flatters any backtest. The diversified sleeves have a few months.
• Spread is the binding cost, and it is not constant. Network fees are negligible; the quoted spread is what every cycle actually pays. It runs between roughly 0.1% and 0.6% round trip depending on the asset, and roughly doubles while the underlying exchange is closed. A ladder whose take-profit is too tight for its asset's spread loses money on every completed trade.
• Quotes come from a counterparty, not a market. These tokens have no order book. Prices are supplied by market makers who may widen, impose minimum sizes, or decline to quote at all - and do. A refused quote delays an exit as readily as an entry.
• A ladder underperforms a strong one-way move. Every cycle caps its own gain, so a ladder on an asset that simply rises earns less than holding it would have. These strategies are built for volatility, and a fast sustained rally is the regime they handle worst.
• The keys live on a server. Autonomous trading rules out a hardware wallet for the trading balance, so that balance is protected by the host rather than by signing hardware. It is kept deliberately small, but it is the one risk here with no market explanation.
• Simulated results are not returns. Every figure here describes what these rules would have produced over specific past windows, with idealised execution. Live results will be worse.