Risks of automated spot trading, named before any order

This note describes risks of the automation Captay wants to build and risks that already exist if you trade or hold coins yourself. It is not a complete list of everything that can go wrong on an exchange. It is not advice to trade or to refrain. The cabinet is closed, so this website currently creates none of the order-path risks below. The text is published so the design is plain before anyone can arm it.

Market risk

Spot prices move. A coin can gap through every level you cared about, in a minute, with no trade at the price you wanted. Holding is not a neutral rest state. A schedule that buys, a grid that accumulates, and a rebalance that sells a winner are all exposed to that movement in different shapes. None of them removes it. Captay does not claim the combination is safer than doing nothing, and does not claim it is safer than holding a single coin.

Automation risk

A rule can be coded wrong, deployed wrong, or left running after you have changed your mind. It can double-send, fail to cancel, or measure a limit with a stale price. Downtime can arrive in the minute the limit should have fired. The halt is designed to stop new risk-taking. It is not designed to be faster than a gap that has already printed. If you need a human to notice first, you need a human awake, not a paragraph.

Exchange risk

The exchange, not Captay, would hold the coins. An outage, a change in API behavior, a frozen withdrawal on their side, or a loss of their solvency sits outside the kill switch. A trade-only key cannot retrieve coins from a venue that will not let anyone retrieve them. Restricting withdrawal protects you from Captay moving coins. It does not protect you from the venue.

Strategy risk

DCA can keep buying a decline until a cap, and the cap can still be large. A grid can fill one side of the ladder and call the other side patience. Rebalance can trim the only position that was working. Overlapping rules can trade against each other and pay two fees to stand still. These are not edge cases pasted on for color. They are the ordinary life of the three rules.

What is not promised

  • No preservation of capital, in nominal or real terms.
  • No result relative to holding, to an index, or to another rule.
  • No assurance that limits will trigger at the print you expected.
  • No assurance that a future cabinet will match these pages.

You can lose money by trading when a cabinet exists. You can also lose money by holding while you wait. Automation does not choose which of those stories you are in. If you cannot afford the loss written into a limit, the limit is already too wide, and the correct state is the one this site is in: no orders.