Investing 202: The Perma-Bear Tax — Deconstructing Negative Carry, Index Mechanics, and the AUD Safe-Haven Paradox

Ask a committed bear why the market has not fallen yet and you get a thesis — valuation, debt, demographics, a yield curve that inverted and has not yet delivered the recession it is supposed to announce. Ask the same investor, several years later, why the thesis was directionally sound and the account is still underwater, and the answers get quieter. That gap is not usually a forecasting failure. It is a structural one. Markets do not compensate an investor for being correct about direction in the abstract. They compensate whoever is positioned correctly for how the instrument they hold actually compounds through time — and the instruments built to express sustained pessimism are, by construction, some of the worst compounding vehicles in finance.

This is not an argument that optimism is virtuous and pessimism is a character flaw. Corrections are real, necessary, and occasionally the healthiest thing that happens to an index in a given cycle. Plenty of investors made money being short in 2008, in the first quarter of 2020, and in 2022. The argument is narrower and more mechanical: a permanent bet against a market-cap-weighted equity index is a bet against the way the instrument is built, as much as against fundamentals. Three separate structural forces work against that position simultaneously — the running cost of holding a short through time, the mathematics of volatility under compounding, and the mechanical self-selection of the index itself. For an Australian investor there is a fourth overlay: a currency relationship that already performs a version of the defensive job the perma-bear believes they are achieving by sitting it out.

Understand all four and the case for permanent bearishness as a portfolio structure — as distinct from a sized, temporary, tactical view — becomes very difficult to defend on the numbers alone.

This piece assumes the fundamentals from the Investing 101 sequence are already in place: dollar-cost averaging, liquidity, market-cap tiers, and the circle of competence that decides what belongs in a plan at all. Investing 202 is where this publication moves from the qualitative filters that decide what you buy to the quantitative mechanics of how a position actually behaves once it is compounding, hedged, or fighting the tide of an index that was not designed to be fought. Expect more arithmetic and less forgiveness for a plan that sounds right in conversation but does not survive contact with how these instruments are actually built.

What this piece is not saying. It is not saying crashes do not happen. It is not saying cash, bonds, gold, or a smaller equity weight have no place. It is not saying every Australian should sit 100% unhedged in a single US index ETF. It is saying that structural short exposure and permanent AUD cash, held as the primary defence, fight forces that have nothing to do with whether the macro thesis is correct.

The Physics of Negative Carry

Carry is the return earned, or the cost paid, simply for holding a position while nothing else happens. A long equity position generally carries positively: dividends accrue, and the position requires no ongoing outlay beyond the capital committed at entry. A structural short runs that relationship in reverse, and the reversal is not a footnote. It is the mechanism by which time itself works against the position, independent of whether the underlying view is ever proven right.

The cost of borrowing conviction

A listed-share short requires borrowing the shares before selling them, and that borrow is not free. Three separate charges accrue for as long as the position is open:

The stock-loan rate. A fee paid to whoever lends the shares, priced by scarcity. Liquid, widely held names and broad index ETFs borrow cheaply — often in the general-collateral band of roughly 0.25–2% a year. Anything genuinely hard to borrow can carry a fee steep enough to erode a position on its own, before the market has moved a single point. The distinction matters. Shorting a crowded small-cap is a different cost structure from shorting the S&P 500 via a liquid ETF or index future. The perma-bear tax on an index short is usually not a spectacular hard-to-borrow special. It is a quiet, compounding combination of a modest borrow fee, manufactured dividends, and the fact that the underlying has a positive long-run drift.

Margin maintenance. Capital posted as collateral is marked to market daily. If the position moves against the short — including on pure noise, unrelated to the eventual thesis — the account must post more collateral or the position is closed involuntarily, at whatever price prevails at that moment. This is not a discretionary risk decision the short-seller controls. It is a contractual one, and it tends to trigger exactly when being forced out is most expensive.

The manufactured dividend. A short-seller owes the lender whatever distribution the borrowed stock pays while the position is open. Shorting a mature, cash-generative business — or an index whose largest constituents are exactly that — means funding the distributions of the companies being bet against, out of the short position itself, on every ex-dividend date, for as long as the position exists. As of mid-September 2026 the S&P 500 trailing dividend yield sat near 1.1%. That is the bill, every year, on a flat tape.

Add the three together and a structural short is not a free-standing bet against a view. It is a position that pays a running toll — borrow fee, manufactured dividend, and the opportunity cost of posted collateral — every day it exists, whether the market moves or not.

A bear can be exactly right about direction and still be mathematically underwater on carry alone if the move arrives late and shallow. The same bear can be paid handsomely if the crash arrives quickly. Carry is not a law that shorts cannot win. It is a clock. The longer the thesis takes, the more the clock charges.

Worked example: the flat-tape invoice

Figures below are illustrative, rounded, and deliberately conservative — a liquid index short, not a hard-to-borrow special.

Hold that position for three years on a tape that goes nowhere. Price change: 0%. Carry at 1.7% a year compounds to a hole of roughly 5% — before brokerage, before any squeeze, before the fact that equities as a class have historically drifted up, not sideways. The thesis can be “the market is expensive” for thirty-six months running and still be consistent with a red account. That is the tax. Treat 1.7% as the visible carry leak on a liquid index short, not the whole economic bill. Posted margin is capital that is not earning the equity risk premium. The size of that second bill depends on the haircut and on cash rates at the time. It is omitted from the 5% three-year hole so the invoice stays conservative and replicable.

Index futures change the packaging, not the core economics of fighting drift. There is no stock loan in the same form, but there is roll, basis, and the same manufactured-dividend-equivalent embedded in the futures curve. A rolled futures short still bets against positive expected equity drift; it does not pay a hard-to-borrow special, and it is not the product this piece is mainly taxing. Daily-reset inverse exchange-traded products avoid the margin call of a prime-broker short and cap loss at capital invested. They introduce a different leak, which is the next section. A purchased put is a third structure again: a known premium, a known maximum loss, no daily reset. That tax belongs in the menu at the end, not in this invoice.

Path dependency: why being right is not enough

Compounding is multiplicative, not additive, and that single fact is where most multi-year bearish product strategies quietly lose money even in a flat or mildly favourable market. The relationship between an investment’s average return and its actual compounded return is captured by the variance-drag approximation:

Geometric return ≈ arithmetic mean return − σ²/2

where σ is the standard deviation of returns. The identity says something specific: volatility itself has a mathematical cost, deducted from the average return to arrive at what an investor actually experiences after compounding. A position with a zero arithmetic expected return still carries a negative expected compounded return the moment volatility is present, and the drag scales with the square of volatility — doubling the volatility roughly quadruples the drag.

The effect is sharpest in daily-reset leveraged and inverse products, which are the vehicles most often used to express a bearish view without stock-loan mechanics. These instruments target a fixed multiple of that day’s return, not the cumulative return over the holding period, which means the sequence of returns — not just their average — determines the outcome.

Consider two consecutive sessions in which an index rises 10% and then falls 10%. The cumulative index move is 1.10 × 0.90 = 0.99, a 1% loss, despite an average daily return of zero. A daily-reset −1× inverse product moves in the opposite sequence — down 10%, then up 10% — and arrives at the identical 0.90 × 1.10 = 0.99, also a 1% loss. The drag is symmetric on that round trip: it punishes the short exposure exactly as it punishes the long one, on the same two days, regardless of which side of the trade the volatility was meant to help.

Two caveats sit next to that classroom example, because the example is often asked to do more work than it can.

First, decay of this kind is a chop tax, not a guaranteed daily leak. In a fast, one-directional decline, a daily-reset inverse product can outperform a static −1× short, because each day’s gain is applied to a larger net asset value as the fund covers into the fall. The product is dangerous in a market that shakes without going anywhere, and useful — briefly — in a market that falls in a straight line. Permanent holders live in the first world. Tactical holders sometimes live in the second.

Second, the same volatility-drag identity applies to the long side. A long-only investor absorbs the same σ²/2 haircut. The difference is what sits underneath it. The long has a structural tailwind: the long-run positive drift of equities, commonly framed as the equity risk premium. A short stacks volatility drag on top of a position that is already betting against an asset class with a positive long-run drift, and stacks that on top of the running carry cost described above. Three negative forces compounding in the same direction, before the bear’s fundamental thesis needs to be correct even once.

The same intuition sits, by analogy rather than by identity, inside the Black–Scholes–Merton option-pricing framework: volatility is not incidental noise. It is a priced structural input. An option’s value rises with volatility because a volatile path carries a cost distinct from the average outcome it produces. A short-seller who treats a multi-year inverse product as if it were a clean −1× on the index over the holding period is, functionally, pricing that path cost at zero when the product is not.

What this section is for. To separate “I think the market will fall” from “I have chosen an instrument that pays rent until it does.” The first is a view. The second is a structure. Only the structure is the subject of this piece.

The AUD/USD Safe-Haven Paradox

Currency markets classify the Australian dollar as a high-beta, procyclical commodity currency: its value moves with global growth expectations, Australia’s terms of trade, and broad risk appetite. The US dollar occupies the opposite structural role — the world’s primary reserve and, in most funding-stress episodes, safe-haven currency. In periods of acute risk aversion, global capital often flows toward USD-denominated and US Treasury assets largely because of that reserve status and liquidity depth, independent of US-specific fundamentals in the moment.

The Reserve Bank of Australia has documented the Australian dollar’s tendency to trade as a liquid, high-beta proxy for global risk sentiment rather than as a defensive asset in its own right. In March 2026, RBA Deputy Governor Andrew Hauser argued that the dollar’s safe-haven status remained “largely intact,” while also recording that it is not a perfect hedge for every risk-off event — including policy shocks such as the April 2025 tariff announcements, where the classic USD bid did not arrive on cue. That double statement is the right one to hold: the pattern is structural and recurring, and it is not a contract.