Copy-trading prediction markets, done right.

Copy-trading prediction markets is the practice of mirroring the positions of large, profitable bettors on Polymarket and Kalshi. The premise is simple: a trader willing to put meaningful capital behind a binary outcome has done analytical work that a retail bettor can borrow against. The execution is harder than the premise suggests. This page walks through the complete framework that survives contact with the market, with a focus on the specific mechanics that separate sustainable copy-trading from the version that blows up inside a quarter.

What copy-trading prediction markets actually means

A prediction market is a venue where binary event contracts trade between zero and one dollar, with the price reflecting the implied probability that the contract resolves YES. The two largest venues today are Polymarket, a non-U.S. platform running on the Polygon blockchain, and Kalshi, a CFTC-regulated designated contract market available to U.S. residents in all 50 states. Both venues attract directional bettors who deploy meaningful capital against specific binary outcomes.

Copy-trading on these venues is the activity of taking the same position as a tracked whale, on the same contract, at a price as close as possible to the whale's executed level. It is not portfolio management, not auto-execution, and not trade-mirroring through a managed service. The copy-trader retains full control of the position, executes manually in their own Polymarket or Kalshi account, and accepts that the copied price will be modestly worse than the whale's executed price.

Why copy-trading prediction markets works (when it does)

Three structural features of prediction markets make whale activity more interpretable than equivalent activity on other trading venues.

The first is the absence of passive flow. No institutional investor holds a portfolio of binary event contracts as an index allocation. Every meaningful position on a prediction-market contract was entered by a trader who made an active decision to deploy capital against a specific binary outcome. The directional intent is unambiguous in a way that it is not on a broad-market exchange.

The second is contract duration. Prediction-market contracts have defined resolution dates, typically within months, and cannot be held indefinitely as a long-term allocation. The short duration means that any whale-sized position must be predicated on a view of the underlying event resolving within the contract's life. Traders who take positions for non-directional reasons, such as market-making or arbitrage, generate distinctive trade patterns that can be filtered out of the directional signal.

The third is market size. Even the largest prediction-market contracts have open interest measured in tens of millions of dollars, which is small relative to traditional financial markets. A whale-sized position is a meaningful fraction of the open interest on a given contract, and the position cannot be entered without showing conviction. The trader who places a six-figure position is doing so visibly, with full knowledge that the trade will be observable and will move the price.

The Rivo copy-trading framework

The framework Rivo applies to surface copy-tradeable whale activity is built around five filters. Each is designed to eliminate a specific failure mode of naive copy-trading.

Filter one: size threshold

Only large opening positions are flagged as whale trades. The threshold filters out retail-scale activity, market-maker rebalancing, and small-scale arbitrage flow. Below the threshold, the population of trades is dominated by behavior that does not carry directional signal.

Filter two: wallet history

Whale trades from wallets without sufficient resolved history are deprioritized. A minimum of twenty resolved positions is a rough floor for confident skill inference. Above that threshold, wallets with positive category-specific PnL are weighted more heavily than wallets without one.

Filter three: contested outcome

Trades on contracts where the YES price is above 0.75 are filtered out as effectively decided. The remaining contested-outcome trades are where the bulk of copy-trading edge lives, because the asymmetric payouts on extreme-price entries reward the trader disproportionately when the call is correct.

Filter four: book depth

Trades on books too thin to support the whale's executed size at the moment of entry are filtered out. The order book must be deep enough that a copy-trader can fill a proportional position within two cents of the whale's executed price. Beyond two cents of slippage, the expected value of the copy-trade has typically degraded enough to be unattractive.

Filter five: AI quality judge

Trades that clear the four deterministic filters above pass to an AI quality judge that evaluates the market title, category, current price, and trade context, and produces a send-or-skip decision. The judge is the final gate before an alert is sent. Approximately one in twenty whale-sized trades clears all five filters and is surfaced as a copy-tradeable signal.

Sizing rules that survive contact with reality

The single largest source of copy-trading failure is sizing the position by reference to the whale's ticket rather than to the copy-trader's own bankroll. A whale who places a $200,000 ticket may be deploying half a percent of working capital. A copy-trader who responds with $2,000 against a $40,000 bankroll is deploying five percent. The same trade, copied at the wrong relative size, becomes a fundamentally different risk profile.

A defensible default is to cap any single copy-trade at two percent of the bankroll, regardless of conviction. Whales lose; the question is whether the copy-trader's capital structure can absorb the losses while compounding the wins. A two-percent cap allows the trader to take roughly fifty independent positions before a string of losses becomes structurally meaningful, which is enough statistical surface area to evaluate the strategy.

How copy-trading prediction markets differs from copy-trading stocks or crypto

Copy-trading on stock or cryptocurrency exchanges is generally a portfolio-level activity: the copy-trader mirrors the entire allocation of a chosen trader and accepts the resulting exposure. Copy-trading prediction markets is position-level. Each copy-trade is an independent decision about a single binary outcome, and the copy-trader can accept or reject each individual signal.

The position-level structure has two important consequences. First, the copy-trader retains analytical control over which trades to take and which to skip; the trader is not committing to every position the whale makes. Second, the copy-trader can size each position independently, applying the two-percent rule per trade rather than allocating a fixed percentage of the portfolio to the whale's full book.

Platforms and pricing

Rivo provides copy-trading signals across both Polymarket and Kalshi, surfaced through three channels: Telegram, web push notification, and an in-app feed. Telegram is the primary surface for U.S.-based copy-traders who use Kalshi. Web push is the primary surface for non-U.S. copy-traders using Polymarket. The in-app feed is the canonical record of all alerts and is available on all platforms. For the detailed mechanics of the Telegram integration, see our Polymarket Telegram bot page. For the broader alerting infrastructure, see our alerts page.

What to read next

For the full copy-trading methodology, including how to evaluate wallets, read entry prices, and execute fills, see our complete copy-trading guide. For the decision framework on when to copy a whale and when to fade them, see when to copy a whale and when to fade them. For the comparative analysis of the two venues, see Polymarket vs Kalshi whales. The full set of free guides is in our learn library.

Frequently asked questions