Why leverage destroys the result even when the forecast is right

Leverage loses on a correct forecast because the forecast and the position answer different questions. A forecast says where the price will end up. A position built on borrowed money does not live at the destination but on the route towards it: it is marked to market every day, its margin requirement is recalculated every day, and the cost of carry is debited every day. It is enough for the route to turn out deeper or longer than expected, and the position will be gone before the forecast comes true. You will still have been right; there will be no result to show for it.

A forecast describes a point, leverage trades the route

Any view on the market consists of a direction, a target and a time frame, and the time frame is usually the weakest element: it is named last and checked least often. Without borrowed funds, a weak time frame costs little — you simply wait. With leverage, time becomes a parameter you pay for, and the depth of the interim drawdown becomes a condition of the position's survival. Leverage turns the qualitative statement "the security is undervalued" into the quantitative commitment "the security will not fall below a certain level before a certain date", even though you never claimed anything of the kind.

This is the source of a common mistake when people review their own trades: they see that the target was reached and treat the close-out demanded by the broker as a technical accident. It is not an accident; it is a different statement of the problem. Leverage does not test your estimate of value — that is what valuation multiples and their limitations are for — but your ability to survive the path.

The arithmetic of a drawdown is asymmetric

Loss and recovery are calculated from different bases. A drawdown reduces capital, and the subsequent gain accrues on that reduced capital, so getting back to the starting point takes a move larger than the drawdown itself. Without borrowed funds this asymmetry is barely noticeable. Leverage scales the drawdown, and therefore scales the gap between the depth of the fall and the size of the rebound required — and that gap grows faster than the drawdown itself.

The practical consequence: a series of leveraged trades in which correctly called directions outnumber the wrong calls can perfectly well end in a loss. The signs matched, but the bases from which the percentages were calculated did not.

The margin requirement changes by itself

The size of the margin depends on the risk assessment of the instrument, and that assessment is not constant. When the market becomes more volatile, risk parameters are tightened: the same position size starts to require more free funds. This happens at the same time as the price moves against you, that is, at the very moment when free funds are at their scarcest.

The result is a loop that runs by itself: the price goes down — the requirement rises — part of the position is closed — the closing weighs on the price — the requirement rises again. Your forecast plays no part in this loop at all.

The cost of carry is charged every day

Borrowed money costs money, and the cost accrues over calendar time, including the days when there is no trading. This turns waiting from free into paid: the position has to earn back the cost of carry before it starts to bring anything to you. The longer the forecast takes to play out, the higher the bar.

The cost of funding itself is tied to the level of interest rates in the economy, so it has to be read together with the regulator's rhetoric: a change of signal alters both the cost of carry and the risk assessment of instruments. For the debt market the link is at its tightest — leverage layered on top of rate sensitivity through duration makes a position in OFZ far more abrupt than its coupon suggests.

The close-out lands on the worst order book

A forced close-out is a market order placed at a moment you did not choose. The risk system chooses it, and chooses it on the basis of price movement, that is, exactly when liquidity drains away and the spread widens. The execution price turns out worse than the price on the screen, and at the boundaries of trading sessions this is compounded by gaps between clearing sessions and by thin evening and morning order books. How settlement and mark-to-market work technically is covered in the article on the central counterparty.

The dividend gap deserves separate mention: a leveraged position passes through the record date with a mechanical drop in price and receives nothing in return if the security sat in the short part of the portfolio or if the close-out came earlier. It is useful to know the dates in advance — the nearest payouts: {{dividend_calendar|limit=5}}, and the full list is in the dividends section.

Several leveraged positions are usually one bet

Leverage spread across different securities looks like diversification and rarely is. Shares of one market, bonds and currency instruments are to a large extent explained by common factors: the interest rate, the exchange rate, risk appetite. In a shock these positions move in step, the margin requirement rises across all of them at the same moment, and what has to be closed is not the worst position but whichever is to hand. An instrument card such as {{instrument:SBER}} will show its own parameters, but it will not show that it is going to move together with the rest of the equity market.

What follows from this

The answer lies not in giving up the forecast but in changing the unit of measurement. Plan not for the expected profit but for the depth you can bear: what drawdown the position will withstand without the risk system stepping in, and how many days it is able to pay for carry. If the answer to the second question is shorter than your horizon, leverage is working against the thesis, not for it. Step 3 matters more here than the first two — check that if all positions move against you simultaneously, free funds are still sufficient.