Copy-trading on prediction markets is the practice of identifying large directional bettors, often called whales, and taking positions that mirror theirs on the same contracts. The premise is straightforward: a trader who is willing to put six or seven figures behind a binary outcome has done work that a retail bettor can borrow against. The execution is harder than the premise suggests. This guide sets out the framework that survives contact with the market, including the wallet evaluation, entry pricing, and sizing rules that separate sustainable copy-trading from the version that blows up inside a quarter.
What qualifies a trader as a whale
The working definition Rivo applies is a single-contract ticket large enough to clear our internal size threshold, placed as an opening position rather than a close or partial exit. The threshold is set high enough to filter out retail behavior, market-maker rebalancing, and small-scale arbitrage flow, and low enough to capture the full population of directional bettors operating at meaningful size. Below that floor the population is dominated by activity that carries no directional information; above it the population thins out to traders who have made a conscious decision to deploy capital against a specific view.
Size alone is not a verdict on skill. A $50,000 ticket from a wallet with no prior resolved history carries less signal than a $15,000 ticket from a wallet that has been on the correct side of dozens of contested markets. The first step in a serious copy-trading workflow is to separate the whales whose track record warrants attention from the whales whose size is the only thing notable about them.
The four-step workflow
Step one: pick a category
Categories on prediction markets are not interchangeable. Politics rewards patience and conviction; sports rewards speed and information edge; macro rewards a familiarity with the underlying data releases; crypto rewards a functional understanding of on-chain mechanics. A trader who is competent in one category is not necessarily competent in another. The first decision is which category to operate in, and the correct answer is the category in which the trader has enough domain knowledge to evaluate whether the whale's call makes sense.
Copy-trading without category knowledge is the most common failure mode. The whale's size tells the copy-trader that the call is worth taking seriously; it does not provide the domain context that determines whether the call is good. Without that context the copy-trader cannot distinguish a sharp entry from an expensive mistake.
Step two: filter wallets by resolved history
Once a category is selected, the next filter is wallet track record within that category. The relevant metric is not raw win rate, which is heavily influenced by entry price, but resolved-trade PnL across at least twenty positions. Twenty resolved trades is a rough floor below which the signal-to-noise on a wallet is poor. Above that threshold, a wallet with a positive PnL across resolved history is meaningfully more likely to be a productive copy-trade target than a wallet without one.
Rivo surfaces resolved-history filters on each tracked whale. The simplest version of this filter is to weight wallets by category-specific resolved PnL and ignore wallets that lack the history to support a real read, regardless of how large their current open position is.
Step three: read the entry price
The price at which a whale enters a market is the most information-dense data point in the trade. A buy at 0.65 on a contract priced near a coinflip is a small directional view with bounded upside. A buy at 0.15 on a contested market is an asymmetric play with significant upside if the call is correct. The two trades look identical in size and direction but reflect very different kinds of conviction.
Sharp money is disproportionately found on the side of the book that the broader market does not want. A whale who consistently enters at extreme prices, holds, and resolves positively is doing analytical work that most participants are not. A whale who consistently enters near the median price of the book and exits early is closer to running an inventory strategy than a directional one. Copy-trading the second kind is harder to make work because the edge per trade is smaller and more dependent on execution.
Step four: pre-commit to a stake size
The single largest source of copy-trading failure is sizing the trade by absolute reference to the whale rather than by reference to the copy-trader's own bankroll. A whale who places a $200,000 ticket may be deploying half a percent of their 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. The rule is to size by percentage of the copy-trader's bankroll, not by ratio to the whale's ticket.
Sizing rules that survive contact with reality
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.
A second rule is to bias toward smaller positions when uncertain. The choice between a $200 ticket and a $500 ticket is almost always resolved correctly by taking the $200. Most copy-trading mistakes share a common shape: position too large for the level of conviction, exit plan too vague, hold duration too dependent on the copy-trader's emotional state. Smaller positions reduce the cost of all three failure modes.
Book depth and slippage
A copy-trade is a different trade from the whale's trade in one critical respect: the whale set the price, and the copy-trader is buying into a book that has already absorbed the whale's flow. The depth of the book at the moment the copy-trader enters determines whether the executed price is meaningfully worse than the whale's executed price.
Rivo surfaces order-book depth in USD on both sides of the market at the moment of the original trade. A useful threshold is to skip any copy-trade where the copy-trader cannot achieve a fill within two cents of the whale's executed price. Beyond two cents of slippage, the expected value of copying the trade has degraded to the point that the trader is paying for the privilege of following someone else's idea.
When not to copy
Several specific situations are reliably wrong to copy, regardless of how good the wallet looks in aggregate.
The first is a trade on a market that resolves within twenty-four hours where the copy-trader cannot monitor the close. Short-dated markets move sharply in the final hours, and a position taken without the ability to manage it through resolution is closer to a lottery ticket than a copy-trade.
The second is a position that, on closer inspection, is a sell rather than a buy. A whale closing a YES position looks similar in the raw tape to a whale opening a NO position; the directional implication is opposite. Rivo labels position deltas where the data supports the inference, but in any case the rule is to trade only on clearly opening flow.
The third is the trade the copy-trader cannot justify in a single sentence. If the copy-trader cannot articulate the thesis the whale is acting on, the copy-trader is copying the action and not the decision. Copy-trading actions without decisions is the same activity as following a tip from a stranger, and produces substantially the same results.
How to use the wins and losses leaderboards
Rivo publishes two public leaderboards. The biggest wins page ranks the largest paper-PnL gains across tracked whales. The biggest losses page ranks the largest paper-PnL losses. Both are useful, but the losses page is more useful for an aspiring copy-trader.
The wins page is a survivorship-biased sample of the trades that paid. Reading it in isolation produces an overoptimistic view of how often whale trades work. The losses page corrects for that bias by surfacing the trades that aged badly, the news chases that mean-reverted, and the conviction positions that resolved against the bettor. The same wallets often appear on both lists. The education comes from understanding why the same trader was sharp on one trade and wrong on another.
The honest framing of the strategy
Copy-trading is not a free signal. It is the practice of running the same trade as a more capitalized, more experienced participant, with a delay between the original execution and the copy and with less context about the underlying view. The realistic expectation is that copy- trading provides a better starting universe of trades than a retail bettor would assemble independently, not that it provides positive expected value on every individual position. The framework above is built to capture the first benefit while controlling for the failure modes that otherwise destroy the strategy.
Frequently asked questions
What is the minimum bankroll for copy-trading prediction markets?
A practical floor is enough capital to take at least twenty-five positions at two percent of the bankroll each, which means at least a few thousand dollars in deployable capital. Below that level the position sizes are small enough that fees and spread become a meaningful drag, and the sample size of copied trades is too small to evaluate the strategy.
How quickly do I need to act after seeing a whale trade?
Faster than most retail traders manage but slower than most retail traders assume. The window in which the price is still close to the whale's executed level is typically between thirty seconds and ten minutes, depending on market liquidity and time of day. A copy-trader who is not in front of the screen when a relevant trade prints is usually better served by setting a limit order at a target price and walking away than by chasing the fill.
Can I copy multiple whales at once?
Yes, and diversifying across several uncorrelated wallets is generally superior to concentrating on one. The practical constraint is that a copy-trader must still evaluate each trade individually rather than blindly mirroring every entry from every tracked whale. Auto-copying without filtering produces a portfolio that looks like the average of the wallets being followed, which is rarely the goal.
What happens if the whale exits before I do?
Frequently the whale will exit a position before resolution, either taking profit or cutting a losing trade. Rivo flags exits when the data supports the inference. A defensible policy is to exit when the whale exits, on the theory that the whale's information advantage at entry is the same advantage at exit. A more aggressive policy is to hold to resolution on the original thesis. Both can work; mixing the two within a single position usually does not.
Is copy-trading legal?
Yes, in the sense that taking the same position as another public participant in a market is not regulated activity. The relevant legal question is whether the platform on which the copy-trade is being executed is legal in the copy-trader's jurisdiction, which is addressed in our guide to the legal status of prediction markets in the United States.