APRLOG / All articles / Where the risk hides in high-APR small-cap products
Where the risk hides in high-APR small-cap products
Scroll down the earn list. The first few stablecoins pay single digits, and then something three digits appears. The first reaction is “that's good value”. The second should be “on what basis?”. This article is about the second reaction.
Directly: a high rate always corresponds to a cost, but the cost is not printed next to it. The three usual sources are project subsidy, intense borrowing demand and liquidity incentives. The matching risks are that the subsidy stops, the demand evaporates, and you cannot sell when you want to.
Whose money is paying this?
Source one: a project subsidy
A new project sets aside tokens to subsidise yield and build a base of holders. Two properties are certain: the allocation is finite and the period is finite. When it runs out or the project stops, the rate falls off a cliff.
More importantly, the subsidy is usually paid in the project's own token, so the “rate” is denominated in an asset of unknown liquidity.
Source two: borrowing demand
An asset being borrowed heavily to short or arbitrage pushes its lending rate up. That is a real market price, with an uncomfortable implication: a lot of people are betting this asset falls, and you are collecting their interest while holding the thing they are betting against.
Source three: liquidity incentives
Programmes designed to deepen a particular pair. These usually come bundled with mechanics like impermanent loss, and the yield is quoted on a more complicated basis.
The arithmetic that matters most
One simple point: 100% APR on a token that halves means you worked for nothing.
With numbers (hypothetical, not about any specific token):
| Case | Units after a year | Price change | Value | Result |
|---|---|---|---|---|
| 100% APR, price flat | 200 | 0% | $200 | Doubled |
| 100% APR, price −50% | 200 | −50% | $100 | Exactly break-even |
| 100% APR, price −70% | 200 | −70% | $60 | Down 40% |
| 30% APR on a stablecoin | 130 | 0% | $130 | Up 30% |
Starting from 100 units at $1. Hypothetical, to show how price offsets yield; not a real product or token.
The second row is the point: a 100% product only needs its underlying to halve for a year of work to net out. For a small-cap token propped up by a subsidy, halving in a year is not an extreme scenario.
Inverted, that gives a useful question: how far does this have to fall before my yield is wiped out? If the answer is a distance you consider entirely plausible, the rate is not as attractive as it looks.
Liquidity: unable to sell when you want to
The most underestimated risk here, because it is invisible until it matters.
A small-cap token's order book can be thin. While you are not selling you never notice; when you want to — usually when bad news arrives and everyone else wants to — your order collapses the price several per cent on the way down. The price on the yield table and the price you can actually transact at may be far apart.
Rough checks: how many pairs does it trade against, how wide is the bid-ask, and how does your intended size compare with daily volume. If your position is an appreciable share of daily volume, you cannot exit without moving the price.
The day the subsidy stops
Subsidised yields all end the same way, and three things happen together:
- The rate reverts to normal and the product loses its appeal;
- Holders who were there for the yield start leaving, adding sell pressure;
- Liquidity providers withdraw too, thinning the book and amplifying the move.
Each reinforces the others, so the best time to leave a subsidised product is generally not after the subsidy ends. That is not a prompt to forecast the date; it is a prompt to decide your exit condition before you enter.
Where these belong
- They are a speculative position, not a savings position. Size accordingly.
- Check what the yield is paid in. If it is the underlying itself, your exposure compounds — see which coin the yield is paid in.
- Work out the break-even fall, write it down, and use it as your reference.
- Do not size up because the rate is high. A high rate signals more uncertainty, not a better opportunity.
A break-even margin you can compute
break-even fall = r ÷ (1 + r), where r is the period return
It comes from a simple identity: after a year you hold (1+r) times the units at (1−d) times the price, and break-even is where the product equals one.
| Annualised | Fall that wipes it out | Reading |
|---|---|---|
| 10% | 9.1% | A ten per cent fall and you worked for nothing |
| 30% | 23.1% | Under a quarter |
| 50% | 33.3% | A third |
| 100% | 50.0% | Half |
| 300% | 75.0% | Three quarters |
Pure arithmetic, assuming a full year held and the yield realised at nominal value. Real outcomes are also affected by liquidity, subsidy withdrawal and the reward asset.
There is a counter-intuitive reading here: the higher the rate the more headroom you appear to have — but a high rate is itself a marker of volatility. A small-cap paying 300% falling 75% in a year is entirely possible; a major asset paying 10% falling 9% is unremarkable.
So the real use of this figure is not picking the largest margin. It is to compute it and then ask honestly whether a fall of that size is a normal event for this particular token.
Subsidised or demand-driven?
| Subsidised | Demand-driven | |
|---|---|---|
| Who pays | The project, in tokens or cash | Borrowers, in real interest |
| How long it lasts | Until the allocation runs out | As long as the demand exists |
| How it ends | Abruptly | Gradually, as demand fades |
| Knock-on when it ends | Often a falling token | Usually just a rate normalising |
| Paid in | Often the project token | Usually the asset or a stablecoin |
| Risk layer | Project + market | Platform |
The two are not equally dangerous. A demand-driven high rate is a temporary market condition; a subsidised one is a commercial arrangement with an expiry date.
Demand-driven high rates are structurally healthy — they reflect genuine borrowing and end by reverting, without adding downward pressure. Stablecoin rates in a hot market are this kind; the mechanism is in why flexible rates move every day.
Three quick tells for which you are looking at: what the yield is paid in (the project's own token means subsidised); whether there is a cap and an end date (caps and countdowns mean subsidised); and how far above comparable assets the rate sits (the excess is roughly the subsidy).
A five-minute pre-subscription check
| # | Check | If it fails |
|---|---|---|
| 1 | What is the yield paid in, and can I sell it? | Yield you cannot realise is not yield |
| 2 | Subsidised or demand-driven? | Subsidised needs an exit condition set now |
| 3 | How wide is the bid-ask? | Wide means exiting costs extra |
| 4 | Compute the break-even fall — is it plausible? | If plausible, the rate is not what it seems |
| 5 | Is the amount within my speculative budget? | If not, don't, whatever the rate |
| 6 | What is my exit condition, and have I written it down? | Unwritten means non-existent |
The first four concern the product; the last two concern you — and those are the ones that decide the outcome.
The same product, held by someone with a size limit and an exit plan and by someone without, produces very different long-run results.
One last thing
This reads like a warning-off, and the intent is not “stay away” but “know what you are buying”.
These products can occupy a place in a portfolio, provided it is the right drawer — the speculative one, not the savings one. The cost of the wrong drawer is not the loss itself but the mindset you bring to it: unwilling to accept a loss, tempted to add after a gain. Both make the outcome worse than the product ever could.
All illustrations here are hypothetical, refer to no specific token or product, and are not investment advice. High-yield products carry correspondingly higher risk to principal; returns are not guaranteed. Assess for yourself and go by the product page's risk section.