What your agent can actually do.
This is a plain accounting of what's real today — not a roadmap, not a pitch. If a feature isn't listed here, assume your agent can't do it yet.
Strategies you can pick
Three engines, selectable when you configure an agent:
Momentum
Ranks your allowed symbols by trailing return and holds the strongest one(s). Rebalances on a schedule you set.
Mean reversion
Buys symbols that have pulled away from their own trend and sells back toward it. The opposite bet from momentum.
None
No automated entries or exits at all — useful if you want the account connected and monitored without the agent trading it.
These are the only strategies an agent can run. We're not going to describe hypothetical ones here — when something new ships, this page gets updated, not before.
Test a configuration before you run it
Before you commit an agent to a strategy, you can run that exact configuration through a walk-forward test: it's optimized on older data and graded only on data it never saw, which is the same discipline a live account faces — the strategy doesn't get to know the future.
It runs in the background (minutes, not seconds) and emails you when it's done. The results page leads with the out-of-sample return — never a cherry-picked in-sample number — alongside a deflated Sharpe ratio with a plain-English reading of what it means, a breakdown of how the strategy did in its best and worst market regimes, and how the result holds up if trading costs turn out to be 2x or 5x what we assumed. Every run is saved, so you can compare configurations side by side later.
If there isn't enough history to test a configuration honestly, the page says so instead of forcing out a number.
Risk limits it enforces itself
These aren't suggestions the strategy can talk itself out of — they're checked on every proposed order, before anything reaches your broker:
Position sizing
Order and position caps are set as a percentage of your account's equity, not a fixed dollar figure, so they scale with your account.
Daily loss halt
If losses on a given day cross the limit you set, the agent stops opening new positions for the rest of that day.
Cash reserve
You can require the agent to always keep a minimum share of the account in cash, untouched.
Symbol allowlist
The agent can only ever trade symbols you've explicitly allowed — nothing outside that list is reachable.
Market hours only
Orders are only placed while the market is open — no after-hours or pre-market surprises.
Options
Available if your broker has approved your account for options trading — we check that approval before letting an agent place an options order, we don't just assume it.
Paper before live
Every agent starts in paper mode: real prices, simulated fills, no real money. Switching an agent to live trading requires a typed confirmation, not just a click — it's a deliberate speed bump for a decision that shouldn't be casual.
Connecting a brokerage
Most brokerages connect through SnapTrade, using your broker's own login — we never see or store your brokerage password. Some brokers instead require a manual API key, which you generate and paste in yourself. Either way, some connections are read-only (we can show the account but not place orders), and you'll see which kind yours is before you subscribe.
Benchmark comparison
Every agent's performance is shown next to what simply holding the same assets would have returned. If the strategy is losing to the benchmark, that's on the page, not hidden.
Protecting what you already hold
If you connect an account that already has positions in it, the agent doesn't get to sell them off on its own. Liquidating an existing holding needs your explicit approval first — you get asked, not informed after the fact.
Plans
Paper is free, for as many agents as you want to run in simulation. Standard ($19/mo) allows up to 2 live agents. Multiple ($49/mo) allows up to 6. Live-trading limits aside, every plan gets the same strategies, risk limits, and walk-forward testing.
What this is, plainly.
ReeseDesk is software you configure and control. It is not investment advice, and we don't recommend what to buy. You choose the symbols, the strategy, and the limits; the agent executes them.
Automated strategies can lose money, and many lose to simply holding a broad index fund. Past results backtested or live don't predict future ones. Only use money you can afford to lose, and start on paper.