Two numbers appear in the same document. The headline promises 16.7%. The body promises 15%. A careless reader sees a rounding error. I see a term sheet that has already told us more than its marketing wanted to. When a security discloses two different yields on the same page, the gap between them is rarely a typo. It is a price. My instinct, trained across a decade of reading contracts that were written to be skimmed, is to treat that gap as the only honest sentence in the release. Everything else โ the 7 million shares, the October 15 pricing date, the business-day dividend cadence โ is context. The discrepancy is the thesis. I do not trust the silence, I audit the code.
Strive is issuing 7 million shares of SATS, a preferred stock priced on October 15, paying a 15% annual dividend distributed every business day. Preferred stock is a hybrid: an equity shell wrapped around a bond-like cash flow. It sits above common equity in the liquidation stack and below debt. That ranking matters more than the coupon, because it determines who eats the loss when the underlying asset โ here, bitcoin โ moves against the balance sheet.
The structure is not novel. Crypto treasury companies have issued preferreds before, typically in the 8% to 12% band. What is novel is the distribution mechanism: roughly 252 payments a year, each about 0.06% of face value. That is a behavioral design, not a yield enhancement. It converts an annual number into a daily drip, raising perceived return while leaving total return untouched. The perceived-versus-actual gap is where retail money usually enters.
What the excerpt does not disclose is the entire skeleton: par value, offering price, liquidation preference, cumulative or non-cumulative status, maturity or redemption rights, convertibility, dividend coverage source, rating, and underwriting arrangement. Six of those eight items decide whether this is an investment or a liability transfer. Their absence is not an oversight. It is a disclosure choice.
Start with the arithmetic, because arithmetic does not negotiate. If the shares are offered at 90% of par โ a 10% original issue discount โ then the current yield on the 15% coupon is 15 รท 0.90 โ 16.67%. That lands almost exactly on the headline figure. A compounding explanation, 252 daily payments at 15% nominal, yields roughly 16.18% effective โ close, but not close enough. The discount explanation fits the number. The compounding explanation does not.
If that reading is correct, the product is being sold below par. Discounted issuance is a concession. A concession means the underwriter had to give ground to clear the book. Read it plainly: demand met resistance. The 16.7% is not generosity. It is the price the seller paid to move inventory.
The coupon is not a feature of the product. It is a measurement of the issuer's cost of capital. A 15% fixed obligation, on an unrated mid-cap treasury vehicle, sits well above investment-grade preferreds near 6% to 8% and high-yield preferreds near 9% to 12%. The spread is not a gift to the buyer. It is the market's estimate of credit, duration, and liquidity risk, compressed into one number.
Now the structural contradiction. On the liability side: a fixed 15% cost, perpetual or long-dated, paid in cash. On the asset side: bitcoin, which produces zero cash flow and moves violently. This is a leverage structure โ high-cost debt financing a non-yielding volatile asset. It works only if bitcoin's long-run return exceeds 15% plus operating costs. That single assumption is the load-bearing wall. Fragility hides in the single point of failure.
I ran a version of this analysis in 2020, modeling oracle delay risk in early Compound pools. The lesson then is the lesson now: the danger is never the headline variable. It is the assumption nobody wrote down. Here the unwritten assumption is that the dividend is funded by operations. If it is funded by new issuance or by selling bitcoin, the structure is a refinancing loop โ new debt servicing old coupons. That is the single point of failure, and the excerpt does not close it.
The missing terms compound the opacity. If the preferred is non-cumulative, the board may legally skip a payment without default. That discretion is the true consideration behind a 15% coupon. If it is perpetual, principal recovery depends entirely on secondary liquidity โ and 7 million shares is not obviously deep. If it carries a call, investors are redeemed in strength and held in weakness: a negative-skew payoff, capped upside, full downside. If it carries no conversion right, the buyer holds credit exposure to bitcoin, not price exposure. The marketing says bitcoin. The contract may say otherwise.
Preferred stock is a trade, not a gift: holders surrender voting control in exchange for payment priority. That means the people buying SATS have no say in the decision that actually determines their outcome โ whether Strive keeps stacking bitcoin or starts selling it to fund coupons. The board declares the dividend. On a non-cumulative instrument, that declaration is a choice, not an obligation. Governance risk here is not about who sits on the board. It is about who does not.
The choice of instrument is itself a signal. Issuing preferred rather than common equity usually means management believes the common is undervalued and will not dilute it at a low price โ or that the market cannot absorb another common offering. Issuing preferred rather than convertibles means management does not want to commit to future conversion dilution and selling pressure. Both readings point the same direction: a bet on stability, and a hedge against the volatility they are simultaneously underwriting.
The excerpt's own framing gives the game away: 'the more telling number is the price buyers agreed to pay for that yield.' That price is missing. Its deletion is the most informative thing in the document.
Here is the contrarian read. Everyone assumes the buyer is harvesting yield and the seller is raising capital. Invert it. The buyer thinks they bought a return; they actually bought credit risk. The seller thinks they sold financing; they actually executed an arbitrage โ betting bitcoin's return beats 15% plus overhead. Both sides believe they hold the alpha. One of them is wrong, and the terms โ not the coupon โ decide which.
This is why the Celsius post-mortem I published in 2022 still applies. The failure mode was never a single bad loan. It was a fixed-yield promise backed by a volatile book, where the coverage ratio was assumed rather than audited. I told that community to exit 80% of volatile positions and hold stablecoins. Most left. The ones who stayed understood that survival is a math problem, not a sentiment. The same test applies here: compute the liquidation preference coverage before you compute the yield.
The only reliable oracle for this instrument is the prospectus, not the press release. Truth is an oracle, not a price feed. Every yield above the risk-free curve is a question the issuer is asking you to answer without the full text. Before the October 15 pricing, the single verifiable act is to read the terms: cumulative or not, perpetual or dated, callable or not, convertible or not, and above all, where the cash for 252 payments a year actually comes from. Proof precedes value; provenance is the only art. The coupon tells you what they will pay. The terms tell you whether they can.