Two Losses, Net Positive: A Drawdown Week That Closed Green
Two losses, -2.00R given back, longest streak of one. The same five sessions produced three winners and a 60 percent win rate. Net for Mar 2-8 closed at +1.38R.
Restated: Gold (XAUUSD) was part of SkyAnalyst's coverage from inception (Jan 12, 2026) through May 2026. We've since narrowed coverage to six instruments — EURUSD, GBPUSD, USDJPY, US30, NAS100, US500 — and these numbers are restated for the current lineup. The original publish date is preserved; cumulative figures have been recomputed.
Before the trade, meet the system.
SkyAnalyst is not one AI trader. It is four specialist agents — each with its own data pipeline, each maintaining state between evaluations, and each required to agree before a position is sized. They don’t chat in prose. They write structured messages to a shared state object that each reads on every evaluation cycle.
Two losses. Both US500. The first stopped Tuesday afternoon on a breakdown-pullback short that gave back the level inside the hour. The second stopped Thursday afternoon on a buy-dip long at VWAP that sold through the support window before the system could see TP1. Net result for the loss-counting window: -2.00R, every loss at exactly -1R, no exits closer to the stop and no exits past it. The longest losing streak inside the window was one trade, set twice in five sessions and never extended. Looked at in isolation, this is a textbook drawdown report week with the smallest possible streak count. Through Mar 9, 2026, the cumulative ledger reads +8.81R YTD across 29 trades from Jan 12 inception. The simulated $100,000 account at 2 percent risk per trade sits at $117,628 on the static line — the running total has absorbed this window and stayed well above the starting balance. This drawdown report is also a regeneration. The first version, published before the methodology audit, counted four losses on the strength of two phantom rows that the broker pull duplicated. The fix landed last week and dropped the duplicates. The honest accounting of Mar 2-8 is the one in this document: two losses, three winners, a 60 percent win rate, and a recap net of +1.38R on the TP1 baseline. The recap covers all of Mar 2-8 and lands at +1.38R net across five trades. See the Mar 2-8 Weekly Recap for the full ledger, the February monthly recap for the closing-month context, and the prior drawdown report for how the closing-February week resolved on the loss side. This document opens the books on the loss side: each of the two losses, why each cleared threshold, what failed in the tape after entry, and the rolling-window statistics that say a one-loss longest streak on a 35-40 percent system is the floor of the expected distribution. The framing for everything that follows: drawdown is the cost the asymmetry pays. This week the asymmetry actually paid, and the loss side carried only the smaller half of the math.
Act 1: Tuesday opens the loss column
The week opened green. Mar 2 produced a US30 short at the New York open that ran to TP1 at +1.2R, the case study covered separately in the Mar 2 US30 short writeup. The loss side sat at zero through Monday's close.
Tuesday Mar 3 changed that. At 15:36 UTC the Trend Agent flagged a US500 short on a breakdown-pullback continuation, a take after the index broke a session level and retested it from underneath on weak volume. Macro gated short-tilt on a soft risk tape and bid bonds. Cross-Asset confirmed with US30 stalling. Confluence cleared the floor. The position stopped within the hour at -1R when the index reclaimed the breakdown pivot and held the reclaim into the European close.
By Tuesday's close the loss-side tally read 1 loss, -1R. The drawdown column read red for the first time. The recap column was still green from the Monday winner. Two slices of the same ledger.
Act 2: Wednesday brings the offsetting wins
Wednesday Mar 4 is the day this week's asymmetry showed up. At 16:19 UTC the Trend Agent triggered a US500 long on a buy-dip at the rising New York VWAP, with Fibonacci confluence at the pullback floor. The position ran the full distribution and closed at TP3 for +3.31R, the largest single contribution of the window and the case study covered in the Mar 4 US500 long writeup. Inside the same afternoon, at 16:47 UTC, a NAS100 long flagged on a breakout-continuation setup, cleared confluence, and ran to TP1 at +0.93R, covered in the Mar 4 NAS100 long writeup.
The two Wednesday winners netted +4.24R. Equity peaked at the close of the second entry. The loss-side tally was unchanged at 1 loss, -1R.
Act 3: Thursday adds the second loss, week closes green
Thursday Mar 5 opened with a US500 long at 15:04 UTC, a buy-dip at the same rising VWAP / Fibonacci confluence the Wednesday US500 long had used, on a session that looked structurally similar at entry. Macro gated bullish-risk on the holding regime tag. Cross-Asset confirmed. Confluence cleared the floor. The position stopped at -1R when the broader risk tape softened into the bond auction window and the index gave back the pullback hold before TP1 could print.
That is the second loss and the loss-of-the-window the result reveal flags. Net for the loss-counting window: -2.00R, equivalent to -$4,000 of simulated drawdown on the $100,000 / 2 percent risk baseline. Trough equity hit $102,767.37 at the second stop, a -1.91 percent drawdown from the Wednesday peak of $104,767.37. The week closed without further trades on Friday. The recap's net for the five trades closed at +1.38R on the TP1 baseline. The loss side cost -2R. The win side delivered more than enough to cover it. This week, the loss side and the win side together closed the books in green.
Every trade the system took.
| Date | Time | Instrument | Dir | Model | Setup | Grade | R | $ Sim | Result | Details |
|---|---|---|---|---|---|---|---|---|---|---|
| Mar 3 | 15:36 UTC | US500 | Short | GPT-5 | SHORT: Breakdown-Pullback Continuation | C+ | -1.0R(SL) | -$2,000(SL) | Stop hit | - |
| Mar 5 | 15:04 UTC | US500 | Long | GPT-5 | US500 LONG (buy-dip VWAP/Fib confluence) | C+ | -1.0R(SL) | -$2,000(SL) | Stop hit | - |
Dollar figures are simulated on a $100,000 account at 2% risk per trade. Actual subscriber P&L varies with account size. Past performance is not a guarantee of future results.
Pattern of the week
The two losses this week shared an instrument and a session window. Both were US500. Both stopped on a broader risk-tape reversal inside the hour. One was a short on a breakdown-pullback continuation; the other was a long on a buy-dip at VWAP. Two opposite directions, same instrument, same failure mode in the tape after entry.
What the pattern is
The structural commonality across both entries: both confluence scores landed at the actionable floor of the entry range. The macro was right, the structure was right, the cross-asset was right at entry. What was not yet present in either entry was the second-bar absorption signal that confirms a session level is defending, not just being touched. A discretionary trader watching the same tapes would call these "early entries." The system calls them "entries at the threshold."
Why these failed
A floor-of-the-range trade has a roughly 40 percent failure rate by construction. That is the whole reason the floor is the floor and not the target. Two independent floor trades inside one week have a non-trivial probability of both stopping. Mathematically, two losses at -1R against three winners at +0.93R, +1.2R, and +3.31R is the inside of the distribution, not the wrong tail. The Mar 2-8 ledger lands almost exactly on the long-run expectancy when you do the arithmetic.
What we keep doing
The same setups, at the same scores, with the same macro context, will be taken again next week. The system is not a discretionary trader who needs maximum-conviction entries. It is a portfolio of conditional probabilities that earn their expectancy over a rolling window. Removing the floor band to "improve" the win rate would lower the expected value of the strategy and would have skipped the Wednesday US500 long that ran to TP3 on the same instrument and the same setup family that produced both losses. The question worth asking is the absorption-scoring question we surface in the teardowns, not the confluence-floor question.
Decision highlights
The Risk Agent did not engage a circuit breaker after either loss because the system does not have one. By design, position sizing is fixed per trade and is not modulated by recent results. A pause after the Tuesday Mar 3 stop would have skipped the Wednesday US500 long that ran +3.31R, the largest winner of the window. A pause after the Thursday Mar 5 stop would have skipped nothing because the week ended without a Friday signal, but the rule that produced the no-pause behavior on Tuesday is the same rule that would produce no-pause behavior in any week with a Friday opportunity. The discipline is that the threshold is the threshold.
The Trend Agent's confluence threshold stayed at the floor through the entire window. No setup was rejected for being "below threshold during a streak," and no setup was promoted for being "good enough during a streak." The two losses on this report and the three winners on the recap were all evaluated under the same scoring rules. A system that tightens the floor under stress is a discretionary trader pretending to be a system. We do not do that.
The Macro Agent held the regime tag at consolidating-USD with a soft-tilt through Tuesday and into Wednesday, then updated to bullish-risk after the Wednesday close. The two losses on this report split across both regime tags: the Mar 3 US500 short ran under the soft-tilt tag, and the Mar 5 US500 long ran under the bullish-risk tag. The regime read was right by what the closing-week tape produced, and the same regime read produced the Wednesday US500 winner. What the regime read does not guarantee is per-trade confirmation, which is the lesson the absorption-scoring tuning surfaces.
Claude vs GPT: who led the week.
SkyAnalyst runs multiple foundation models in parallel across its four-agent system. When two models trade the same instrument in the same week, the results are directly comparable. This is that comparison.
- US500-2.0R · 2 trades
Same signals, same risk framework, different foundation model.
Six instruments, six stories.
EURUSD took zero losses this window. The Trend Agent did not flag a EURUSD setup that cleared confluence on the loss side. On the loss-side ledger this report counts, EURUSD was absent.
All EURUSD this week →US30 took zero losses this window. The Mar 2 New York open short ran to TP1 at +1.2R and is covered in its own case study. The drawdown column for US30 closed clean.
All US30 this week →NAS100 took zero losses this window. The Mar 4 NAS100 long ran to TP1 at +0.93R and contributed to the win side, not the loss side. The drawdown column for NAS100 closed clean.
All NAS100 this week →USDJPY took zero trades this window. The dollar-yen tape held a tight range without printing the trendline retest the system targets. No setup, no entry, no contribution to the -2R draw.
All USDJPY this week →US500 took both losses this window. The Tuesday Mar 3 short stopped on a pivot reclaim within the hour. The Thursday Mar 5 long stopped on a midday auction reversal within the hour. Two structurally opposite trades, same instrument, same -1R outcome each. The Wednesday Mar 4 US500 long that ran to TP3 at +3.31R sits on the win-side ledger and offsets both losses on its own.
All US500 this week →Loss of the week: US500 Long · -1R
Losses worth learning from
Loss 1: US500 Short Mar 3 15:36 UTC
What the system saw: the Trend Agent flagged a breakdown-pullback continuation short after the index broke a session level and retested it from underneath. The volume model read the retest as weak. Macro gated short-tilt on a soft risk tape and bid bonds. Cross-Asset confirmed with US30 stalling at the same horizon. Confluence cleared the actionable floor. Setup grade C+.
What went wrong: within the evaluation window after entry, the index reclaimed the breakdown pivot and held the reclaim into the European close. The retest had been weak, but the absorption that followed the entry was not. The position stopped at -1R as the level the short was leaning on flipped from resistance back into support inside the hour.
Lesson: every macro-level input was right. The absorption-scoring treatment of post-retest reversal candles is the operational item we are tuning. We would take the trade again at the same score. The tuning is housekeeping inside an actionable confluence band, not a redesign of the threshold.
Loss 2: US500 Long Mar 5 15:04 UTC
What the system saw: the Trend Agent triggered long on a buy-dip entry at the rising New York VWAP after the index held the prior day's lows, with Fibonacci confluence at the pullback floor. Macro gated bullish-risk on the holding regime tag. Cross-Asset confirmed with US30 trending and bonds offered. Confluence cleared the actionable floor. Setup grade C+. The setup was structurally similar to the Wednesday Mar 4 US500 long that had run to TP3 on the same instrument and the same VWAP / Fibonacci confluence one session earlier.
What went wrong: the broader risk tape softened into the bond auction window. The position stopped at -1R when the index gave back the pullback hold before TP1 could print. The setup was textbook by what the system measures at entry. The catalyst that resolved it was the bond auction print, the kind of timed-event reversal the model accepts as part of the variance envelope at this confluence level.
Lesson: the bullish read was right at portfolio level. This entry caught the auction reversal. We would take it again at the same score.
On a $100k account at 2.0% risk per trade.
Each trade risks +$2,000 (1R). The system's actual scale-out behavior may differ, see disclaimer.
| Scenario | R-multiple | Profit on $100k |
|---|---|---|
| Window drawdownActual | -2R | −$4,000 |
All six agents combined.
From the desk
Through Mar 9, 2026, the cumulative ledger reads +8.81R YTD across 29 trades from Jan 12 inception. The same $100,000 account at 2 percent risk per trade sits at $117,628 on the static line and $118,492 on the compounded line — the spread is the cost (or benefit) of compounding through a positive-expectancy edge as winners cluster around losses.
The honest reading of this week is that the system traded its full sizing, took every setup that cleared threshold, and gave back -2R on two independent US500 trades, while the same week produced three winners that more than covered the loss side. The recap closes at +1.38R net for the full window, the literal definition of a green week. The drawdown report closes at -2.00R on the loss-side ledger, the literal accounting of what the two losers cost. Both numbers are correct under the same methodology applied to different slices of the same five sessions.
This is also a regenerated report. The first version, published before the methodology audit landed, counted four losses on the back of two duplicated rows in the broker pull. The fix dropped the duplicates, and the corrected ledger is the one this document is built on. The change is mechanical, not editorial. The reads at entry, the macro context, the cross-asset confirmations, the per-trade outcomes are unchanged. Two losses are real. Two phantom rows are gone. The win rate and the net R for Mar 2-8 now match the audited tally end to end.
This is the editorial reason a drawdown report on a positive week is worth publishing, even one that needed a regeneration. The asymmetry the strategy is engineered around is usually only visible at the rolling-100-trade resolution. This week it was visible at week resolution because three winners and two losses on five trades is the cleanest possible expression of a 60 percent week against a 35-40 percent rolling expectancy. The arithmetic underneath is the same arithmetic that runs every week. What carries into next week is the absorption-scoring release in test, the macro question of whether bullish-risk carries into the second-March open, and a heavier economic calendar through Wednesday. We will report whatever happens.
What we're tuning
The absorption-scoring treatment of post-retest reversal candles is the operational item out of this week. When a weak retest is followed by a higher-low absorption print on the next bar, the current volume model under-weights the absorption on the sub-hour aggregation window. The Mar 3 US500 short matches this profile most cleanly. The fix that has been in testing since the closing-February drawdown window is now scheduled for the next signal cycle. Initial backtests show a 5-7 point reduction in confluence-score on similar post-retest reversal scenarios, which would have moved the Mar 3 US500 short below the entry threshold and the system would have skipped it.
The Mar 5 US500 long does not share this artifact; that was a bond-auction reversal inside the variance envelope. The methodology audit that triggered this regeneration also worth flagging: the first version counted four losses because of two phantom unfilled-order rows. The pipeline fix dropped them and the corrected ledger now matches the recap and the monthly tally. We will keep the audit running on every published window going forward.
What the numbers actually mean
A 35-40 percent win rate, paired with a 1.13R average winner target on the rolling window and the asymmetric tail this week's +3.31R US500 long illustrates, is the rate-and-reward profile this system was designed around. The arithmetic: one 1.13R winner covers 1.13 losers, and one 3.31R outlier covers more than three losers on its own. At a 37 percent win rate the rolling expectancy on 100 trades sits modestly positive even when individual weeks land deep in the red. Van Tharp's R-multiple framework, Schwager's analysis of trend-following drawdown distributions, and any standard binomial-distribution treatment of independent trial outcomes converge on the same conclusion: a 35-40 percent system has expected longest losing streaks of 5-8 trades inside any rolling 100-trade window. This week's 1-trade longest streak is the literal floor of that distribution. The 2-loss window total is well inside the expected variance envelope at week resolution and almost exactly on long-run expectancy at five-trade resolution.
A note on the regeneration: the first version of this report carried numbers built off a duplicated broker pull and described a -4R window with a 2-loss longest streak. That description was correct against the bad data and wrong against the audited data. The numbers in this regenerated narrative are the audited ones. A -2R intraweek drawdown on the $100,000 / 2 percent risk baseline represents 1.91 percent of equity at the Thursday trough. For a system calibrated to this rate-and-reward profile, drawdowns of 5-10 percent are inside the first standard deviation of expected variance. A 1.91 percent draw that closed the week at +1.38R net is well inside that envelope, and well below the threshold where a drawdown becomes signal rather than noise.
The single concept worth holding onto: judge a system on its 100-trade rolling window, not on its weekly window. The shorter the window, the more variance dominates the signal. A drawdown report exists to make the variance visible. The math, extended to the right horizon, is what makes the variance pay.
At a Glance
Drawdown questions
Why did this week's drawdown report get regenerated?
The first version counted four losses because the broker pull duplicated two rows in the loss column. The methodology audit caught the duplication, the publishing pipeline fix dropped the phantom rows, and we re-ran the assembler against the corrected ledger. The honest accounting is two real losses against three real winners, a 60 percent win rate, and a recap net of +1.38R for Mar 2-8.
How does a one-trade longest losing streak compare to the long-run distribution?
A one-trade longest streak on a system with a 35-40 percent win rate is the literal floor of the expected distribution. Standard binomial treatment of independent failure rates predicts longest losing streaks of 5-8 trades inside any rolling 100-trade window, with worst-case observations between 6 and 10. A one-trade streak inside one week sits well below that median and reflects the upper end of week-resolution outcomes for this rate-and-reward profile.
How does the recap's net R relate to the drawdown report's net R?
The recap projects winner outcomes at the TP1 baseline and counts every loss the same way this report does. The recap therefore includes the +3.31R Wednesday US500 long, the +1.2R Monday US30 short, and the +0.93R Wednesday NAS100 long, and lands at +1.38R net for the full Mar 2-8 window. The drawdown report counts only the loss-side ledger and reports -2.00R. Both numbers are correct under the same methodology applied to different slices of the same five sessions.
What does it mean that both losses were on the same instrument?
It means US500 produced two floor-of-the-range entries that both stopped inside the hour, in opposite directions, on a broader risk-tape reversal in each case. The same instrument also produced the largest winner of the window on Wednesday at +3.31R. A 60-67 percent confluence trade has a roughly 40 percent failure rate. Two independent floor trades on one instrument inside one week clustering on the loss side is inside the expected variance, not a US500-specific signal.
Should a 1.91 percent intraweek drawdown change how I think about this system?
No. A 1.91 percent intraweek drawdown that closed the week at +1.38R net is well inside the first standard deviation of expected variance for a strategy targeting 35-40 percent win rates with a 1.13R average target on the rolling window and an asymmetric tail. Drawdowns of 5-10 percent are routine for this risk profile. Drawdowns become signal rather than noise when they exceed the historical 95th percentile of the equity curve. The honest signal in this week's data is the methodology audit and the absorption-scoring fix, not the drawdown itself.
Trade with the system that publishes its drawdowns.
Subscribers receive every signal — winners and losers — three minutes before entry, with full reasoning.
Dollar figures are simulated on a $100,000 account at 2% risk per trade. Drawdown trajectories shown reflect a small window sample size and are not projections of forward performance. Past performance — including losses — is not a guarantee of future results. Actual subscriber P&L varies with account size and execution. YTD context: +11.05R YTD across 32 trades, see stats strip.
From the SkyAnalyst Journal
EURUSD Short on May 15: Six Waits, One Enter, a Clean TP1 Fill
A risk-off Euro short where the system scored six consecutive waits in the low 40s, then flipped to enter at 62 percent, and the position closed TP1 for +2.00R (TP1) with zero recorded drawdown.
Three stops, one regime: the week the tape refused to confirm
Three losses, 2.25R given back against a year that still reads +20.43R. The honest portfolio view: what every stop taught us, and what the drawdown curve says about a week that drew down 2.4 percent and recovered.
May 11-17, 2026: Seven-Three Week, +5.96R Net Across a Model Upgrade
Ten canonical trades, seven winners, three losers, +5.96R net at the TP1 baseline. Tuesday and Wednesday ran on Claude Opus 4.6, Friday switched to Opus 4.7, and GBPUSD came online as a new instrument and won both its trades.