SALCEDO.
connecting PAPER ACCOUNT
A live record, not a pitch

A real trading edge, run in public, on $1,000 of paper money.

Salcedo runs one strategy: trend-filtered short-horizon mean reversion on liquid US equities, holding one to five days. Every rule is published below. Every trade lands in the ledger automatically. The drawdown is printed as large as the return, because that is the half of trading every other site hides.

Account equity
—
Total return
—
Max drawdown
—
Win rate
—
Awaiting first data load.
01 · Ledger

The live account

Equity, daily close

Open positions

SymbolQtyEntryLastOpen P/L
No open positions.

Closed round trips, most recent

SymbolQtyEntryExitP/L
No closed trades yet.
02 · Rulebook

The method, in full

R1
Universe
Liquid US large caps. Price above $10, twenty-day average dollar volume above $25m. No microcaps, no story stocks.
R2
Trend filter
Longs only in names closing above their 200-day simple moving average. Falling knives in downtrends are not bought, however oversold.
R3
Entry
Two-period RSI closes below 10 following at least two consecutive down closes. The pullback is bought at the next session's open.
R4
Sizing
Five equal slots of 20% of equity, fractional shares. On a $1,000 account that is roughly $200 a position. No averaging down, no doubling up.
R5
Exit
Close above the 5-day moving average, or two-period RSI above 70, or five sessions elapsed. Whichever arrives first. The time stop is not optional.
R6
Event guard
No new entries within 48 hours of an FOMC decision, or within five sessions of the name's earnings date. Mean reversion has no edge against a scheduled binary.
R7
Halt
A 10% peak-to-trough drawdown on the account stops all new entries pending a written review, published here.
R8
No hands
Orders are placed by the engine and tagged at fill. Nothing on this page is a manual trade, and no losing trade will be quietly deleted.

These parameters are published before the fact and any change to them will be logged on this page with a date. A rule that can be bent after a losing week is not a rule, it is a mood.

03 · Honesty

The arithmetic nobody advertises

Pick two

High win rate. Large average win. Smooth equity curve.

No method holds all three, because anything that did would be arbitraged out of existence. This system picks a high win rate and a smoother curve, and pays for them with small average wins and occasional sharp losses when a pullback keeps falling. That bill always arrives. The only choice is which currency you pay it in.

Losing streaks are scheduled

At a 65% win rate over 300 trades, a five-loss run is near certain. A six-loss run is a coin flip.

That is not bad luck, it is arithmetic that arrives on time. Position sizing decides whether the streak is a dent or the end of the account. It is why R4 and R7 exist, and why this account risks slots of 20% notional rather than leverage.

The best ever barely won

The most profitable fund in history won about 50.75% of its trades.

Renaissance's Medallion fund earned its famously smooth returns from millions of small bets and industrial diversification, not per-trade certainty. Smoothness is a property of bet count. A $1,000 account has a tiny bet count, so this equity curve will be lumpy even if the edge is real. Judge it after 100 trades, not after a good fortnight.

What $1,000 to $100,000 actually costs

Compounding rateTime to 100×
20% a year (excellent)25 years
39% a year (best fund in history, net)14 years
10% a month (fantasy land)4 years

This site is not attempting any of those. It is attempting something rarer: proving in public, with timestamps, that a small edge exists at all.

Sources for the figures above: Medallion's long-run net returns and trade statistics are documented in Gregory Zuckerman's “The Man Who Solved the Market” and Bradford Cornell's published analysis of the fund. Streak probabilities follow directly from the binomial distribution.