PMTS Monthly Performance Deep-Dive: October 2026 — Wins, Losses and the Equity Gap
Every month we publish a long-form review of what the PMTS algorithm actually did, taken directly from the statistics database that feeds the platform dashboard. This is the November 2026 edition, covering the period to 31 October 2026. It is a performance deep-dive in the literal sense: we show the reference account that has been tracked since July 2025, the monthly results of the other tracked accounts, and one uncomfortable chart that most marketing pages would leave out.
If you only read one paragraph, read this one. The reference account (internal ID 26) is up 25.32% since inception with a shallow realized drawdown. The largest tracked account (internal ID 1) lost money in October on a realized basis and carries a very large gap between balance and equity in the platform snapshots. A system can look excellent on one set of metrics and fragile on another, and a serious reader needs both. The rest of this article explains how to read the numbers, what they say, and what they cannot say.
1. Executive summary: the headline numbers
| Metric | Value | Source / note |
|---|---|---|
| Reference account (ID 26) balance | 62,659.24 | Initial deposit 50,000; no withdrawals |
| Net profit since inception | +12,659.24 (+25.32%) | First trade 21 July 2025, last sync 1 November 2026 |
| Closed trades | 137 (116 wins, 21 losses) | Win rate 84.67% |
| Profit factor | 4.39 | Gross profit 16,651.07 / gross loss 3,792.19 |
| Maximum drawdown (realized) | 1,181.97 (1.995%) | Closed-trade basis |
| Largest single loss | -1,147.12 | About 8 average wins |
| October 2026, account 1 | -131,022.31 (-0.74%) | 6 trades, 4 wins, profit factor 0.31 |
| October 2026, account 24 (partial) | +16,169.60 (+1.52%) | 14 trades, 14 wins; row last updated 13 October |
| Account 1 equity vs balance, 1 Nov | 5.92M vs 17.67M | Floating result of -11.76M in the latest snapshot |
Chart description: a horizontal bar chart with one bar per tracked account showing October 2026 realized return in percent. Account 24 would show a short green bar to the right (+1.52%), account 1 a short red bar to the left (-0.74%). The point of the chart is that the monthly spread across accounts is small compared with the equity gap discussed in section 6.
2. How to read these numbers (and what we left out)
The figures in this article come from the statistics tables that the PMTS platform fills from MetaTrader 5 account data: an overall-statistics table, a monthly table, a daily table and account snapshots. Before comparing anything, four definitions matter, because they change the answer.
Realized versus floating
The monthly and daily tables are built from closed deals. A position that is still open and deeply underwater does not appear in them until it is closed. Account snapshots, by contrast, record balance and equity at a point in time, and equity includes floating profit or loss. A track record that only reports closed trades can therefore look calm while open exposure is accumulating pain. We show both.
Swap is separate from the monthly profit field
In the monthly table, the profit field excludes swap, the overnight financing charge. We verified this on account 1: the October balance moved from 17,808,755.28 to 17,664,067.11, a change of -144,688.17, which equals the reported profit of -131,022.31 plus swap of -13,665.86. Where swap matters we state the net figure.
Balance changes include deposits and withdrawals
Balance-to-balance comparisons are polluted by cash movements. Account 24, for example, shows a September profit of +29,287.95 and a swap credit of +1,646.61, yet its balance fell from 1,065,992.87 to 1,061,342.78. The difference of roughly 35,600 is consistent with money leaving the account, not with trading losses. We therefore use the profit field, not the balance delta, to describe trading results.
Data freshness
Not every row is updated on the same schedule. The October row for account 24 was last updated on 13 October, so it covers about half the month at best. We flag it as partial and do not annualize it or compare it directly with full-month rows.
| Source | What it contains | Limitation |
|---|---|---|
| Overall statistics (account 26) | Cumulative trade statistics since inception | Some fields (holding time, consecutive streaks, Sortino) are not populated |
| Monthly statistics | One row per account per month, closed deals | Excludes floating P/L; profit field excludes swap |
| Daily statistics | One row per account per trading day | Only days with closed trades appear |
| Account snapshot | Balance, equity, margin at a timestamp | Point-in-time; says nothing about the path |
| 30-day deal summary | Sum of all deal rows | Mixes accounts, currencies and cash operations; not a performance measure |
That last row deserves a sentence. The platform API also returns a raw 30-day deal summary: 207 deal rows and a net sum of -1,050,154.55. It adds together every deal on every tracked account, including balance operations and accounts denominated in different currencies, so it is not a performance figure and we do not use it. We mention it only because anyone reading the raw API output would otherwise wonder why it is absent.
3. The reference account: 137 trades, 15 months
Account 26 was funded with 50,000 on its first trade date of 21 July 2025 and has had no withdrawals. It is the cleanest series we have because it has a single deposit, no cash movements to adjust for and a continuous history. Its profile is a textbook example of a high-win-rate, low-payoff system, and understanding that profile is more useful than celebrating the win rate.
Cumulative statistics
| Statistic | Value |
|---|---|
| Initial deposit | 50,000.00 |
| Current balance | 62,659.24 |
| Gross profit | 16,651.07 |
| Gross loss | -3,792.19 |
| Swap (financing) | -199.63 |
| Net profit | 12,659.24 |
| Total return | 25.32% |
| Total trades | 137 |
| Winning / losing trades | 116 / 21 |
| Win rate | 84.67% |
| Profit factor | 4.39 |
| Expected payoff per trade (net) | 92.40 |
| Total volume | 14.88 lots |
| Average volume per trade | 0.109 lots |
| Largest win / largest loss | +896.71 / -1,147.12 |
| Maximum drawdown (closed-trade basis) | 1,181.97 (1.995%) |
| Trading days | 155 |
Long versus short
| Direction | Trades | Wins | Win rate |
|---|---|---|---|
| Long | 96 | 82 | 85.42% |
| Short | 41 | 34 | 82.93% |
| Total | 137 | 116 | 84.67% |
The system is long-biased: 70% of trades were buys. That is consistent with a gold market that spent much of 2025 and 2026 in a structural uptrend, and it tells you the results are partly a function of regime. A long-biased system reports its best numbers when the underlying trends higher. The short side is not much worse in win rate, which is encouraging, but 41 trades is a small sample.
The payoff structure
Dividing gross profit and gross loss by the number of winners and losers gives an average win of 143.54 and an average loss of 180.58. The payoff ratio is therefore 0.79: the average loser is about 26% larger than the average winner. That is normal for a strategy that takes frequent small profits and gives positions room to recover.
| Quantity | Value | Meaning |
|---|---|---|
| Average win | 143.54 | Gross profit / 116 |
| Average loss | 180.58 | Gross loss / 21 |
| Payoff ratio | 0.79 | Average win divided by average loss |
| Break-even win rate | 55.7% | 1 / (1 + payoff ratio) |
| Actual win rate | 84.7% | Margin of safety of about 29 points |
| Largest loss in average wins | 8.0 | 1,147.12 / 143.54 |
| Swap as share of net profit | 1.6% | 199.63 / 12,659.24 |
The break-even calculation is the most useful number in the table. With a payoff ratio of 0.79, the strategy only needs to win 55.7% of the time to break even before costs. It has been winning 84.7% of the time. That 29-point cushion is the real margin of safety, and the question for any forward-looking reader is how fast it would erode if the market regime changed. If the win rate fell to 70%, the same payoff structure would still be profitable. If it fell to 55%, it would not.
The flip side is tail risk. The largest loss, 1,147.12, equals roughly eight average wins. One bad trade erases a week or more of ordinary profits. That is exactly why drawdown control, the subject of our September risk-management guide, is not optional for this style of system.
A note on the Sharpe ratio
The database reports a Sharpe ratio of 7.42 for this account. We do not recommend citing it. A Sharpe ratio above 3 over a year is rare even for the best funds, and a value of 7 almost always indicates that the calculation uses per-trade returns on a short, smooth series rather than annualized returns on a consistent time grid. We show the figure for completeness and treat profit factor, drawdown and the payoff structure as the more reliable descriptors until the calculation is audited.
4. Monthly breakdown across tracked accounts
The monthly table returns the six most recent rows. Together they cover four accounts and three months. We list every row, including the losing ones.
| Month | Account | Start balance | Profit (excl. swap) | Return | Trades | Win rate | Profit factor | Swap |
|---|---|---|---|---|---|---|---|---|
| Aug 2026 | 3 | 61,046.27 | -2,262.45 | -3.71% | 4 | 25.00% | 0.26 | n/a |
| Sep 2026 | 3 | 58,388.09 | -2,463.18 | -4.22% | 3 | 0.00% | 0.00 | -3.15 |
| Sep 2026 | 1 | 17,710,522.67 | +104,797.84 | +0.59% | 6 | 100.00% | n/a (no losses) | -6,565.23 |
| Sep 2026 | 24 | 1,065,992.87 | +29,287.95 | +2.75% | 41 | 95.12% | 59.52 | +1,646.61 |
| Oct 2026 | 1 | 17,808,755.28 | -131,022.31 | -0.74% | 6 | 66.67% | 0.31 | -13,665.86 |
| Oct 2026 (partial) | 24 | 1,061,342.78 | +16,169.60 | +1.52% | 14 | 100.00% | n/a (no losses) | +1,014.41 |
Chart description: a grouped column chart with months on the horizontal axis (Aug, Sep, Oct) and monthly return percent on the vertical axis, one colour per account. Account 3 would appear as two red columns at roughly -3.7% and -4.2%, account 24 as two green columns at +2.75% and +1.52%, account 1 as a small green column in September and a small red one in October. The visual message is dispersion: the same algorithm family produced results from -4.2% to +2.7% in a single month depending on the account.
What the table tells us
Three observations stand out. First, the results are dispersed. In September, account 24 returned +2.75% on 41 trades, account 1 returned +0.59% on six trades, and account 3 returned -4.22% on three trades. These accounts differ in size, broker, trade frequency and, as far as we can tell from the trade counts, in how aggressively they are configured. Dispersion across accounts is information, not noise: it tells you that results depend on configuration and account conditions as well as on the signal.
Second, account 3 lost money in both months it appears, with a combined realized loss of 4,725.63 and a single winning trade in seven. It is the smallest sample in the table, but it is also the clearest failure, and we include it for the same reason we included it in September: a monthly report that shows only the green rows is an advertisement, not a report.
Third, trade counts are very different. Account 24 trades often (41 trades in September), while account 1 trades rarely (six trades per month) but with enormous size, totalling 60 lots in September and 100 lots in October. Low frequency with large size produces lumpy results, which is what October shows.
5. October in detail: the two trades that decided the month
Account 1 made six closed trades in October. Four won for a gross profit of 58,039.29, and two lost for a gross loss of 189,061.60. That is a win rate of 66.67% and a profit factor of 0.31, and it produced a realized loss of 131,022.31, or -0.74% of the opening balance. After swap of -13,665.86, the net change was -144,688.17, or -0.81%.
| Date | Trades | Daily result | Daily return | Volume (lots) | Comment |
|---|---|---|---|---|---|
| 7 Oct | 1 | +16,738.61 | +0.09% | 10 | Ordinary winner |
| 14 Oct | 2 | -189,061.60 | -1.06% | 30 | Two losers; the largest was -137,870.87 |
| 15 Oct | 1 | +3,905.03 | +0.02% | 10 | Small recovery |
| 23 Oct | 1 | +13,292.29 | +0.08% | 25 | Winner |
| 26 Oct | 1 | +24,103.36 | +0.14% | 25 | Winner |
Chart description: a daily result column chart for October on account 1, with four small green columns and one tall red column on 14 October that is roughly eight times the size of the largest green column. A cumulative line overlaid on it would drop sharply on the 14th and then climb slowly through the rest of the month without recovering the loss.
The arithmetic is the same asymmetry we saw on the reference account, only more extreme. On 14 October, a single day with two losing trades erased about 3.3 times the profit that the four winning trades earned across the whole month. The largest single loss, 137,870.87, was 0.77% of the account balance. In isolation, that is a modest risk per trade. The problem is not that the stop-loss logic failed; it is that a strategy with a payoff ratio below one needs a high win rate, and with only six trades a month there are not enough trades for the win rate to average out. Two losses in six trades is statistically unremarkable. It is also enough to turn a month negative.
For readers who care about method: low-frequency, large-size trading has a high variance of monthly outcomes. September (six wins, +0.59%) and October (four wins, -0.74%) are the same distribution sampled twice. Judging the system on either month alone would be a mistake in both directions.
6. The chart most reports leave out: balance versus equity
Everything above is based on closed trades. The account snapshot and the daily table, however, also record equity, which includes open floating profit and loss. For account 1 the gap between balance and equity is large and persistent.
| Date | Balance | Equity | Equity minus balance | Gap vs balance |
|---|---|---|---|---|
| 14 Oct | 17,624,510.51 | 10,512,890.97 | -7,111,619.54 | -40.4% |
| 15 Oct | 17,626,952.63 | 11,772,665.85 | -5,854,286.78 | -33.2% |
| 23 Oct | 17,640,244.92 | 8,669,591.64 | -8,970,653.28 | -50.9% |
| 26 Oct | 17,664,067.11 | 8,907,056.35 | -8,757,010.76 | -49.6% |
| 1 Nov (09:09 snapshot) | 17,672,877.18 | 5,915,011.70 | -11,757,865.48 | -66.5% |
Chart description: a two-line chart with balance as a nearly flat line near 17.6 million and equity as a second line oscillating between roughly 12 million and 6 million through October, with the area between the lines shaded. The shaded area is the floating loss. Its widening on 1 November is the main fact of this section.
The latest snapshot shows 12 open positions, margin used of 437,572.87 and a margin level of about 1,352%. The margin level is the more reassuring number: with equity more than thirteen times the margin requirement, there is no near-term margin-call risk at current prices. The equity gap itself is not reassuring. A floating loss equal to two-thirds of the balance means that the realized statistics in the monthly table understate the risk the account has actually been carrying.
We want to be precise about what we know and do not know. The snapshot is a single data point. The list of open positions returned by the same API is stale: it contains rows from January that are flagged as closed, so we cannot reconstruct from this feed which positions produce the floating loss or how old they are. The equity field on account 24 in the daily table (for example 10,606.63 against a balance of 1,074,041.74 on 7 October) is clearly inconsistent with its balance and almost certainly reflects a reporting artefact, so we have excluded it. We cannot say whether the account 1 gap is a genuine floating loss, a data-feed problem or a feature of how the account is used. What we can say is that the gap exists in the platform's own data, and that anyone who evaluates PMTS on closed-trade statistics alone is evaluating only half of the picture.
Why this matters for an institutional audience
Institutional allocators look at maximum adverse excursion, time under water and equity drawdown, not at closed-trade win rates. A strategy that holds positions through large adverse moves can post a high win rate and a small realized drawdown while the equity curve is deep underwater. Our own reference account shows a closed-trade drawdown of 1.995%, which is genuinely low, but a closed-trade drawdown is not an equity drawdown. Until the platform publishes equity-based drawdown for every tracked account, the safe reading of our public statistics is that they describe realized performance and that equity risk needs to be assessed separately.
7. Short-term context: the last seven days
The API's seven-day summary, 25 October to 1 November, shows 28 deal rows, 26 winning deals, no losing deals and a net result of +124,279.12, a win rate of 92.86% when the two non-winning rows are counted. This figure aggregates all accounts and includes zero-profit rows, so we treat it as indicative only. The direction is consistent with the daily rows: account 1 closed winning trades of +13,292.29 on 23 October and +24,103.36 on 26 October, which is the only trading activity we can verify on the large account in that window.
A good week does not change the equity gap above. We mention it because the contrast is instructive: realized results can improve every week while floating exposure keeps growing.
8. Lessons from the data
Four lessons follow from the numbers, and they apply to any systematic strategy, not only to PMTS.
| Lesson | Evidence | Implication |
|---|---|---|
| High win rate is not safety | 84.7% win rate, but largest loss equals 8 average wins | Judge systems by tail loss and payoff ratio, not win rate |
| Small samples mislead | Account 1: +0.59% in September, -0.74% in October on 6 trades each | Do not extrapolate a month with fewer than 30 trades |
| Realized and equity views diverge | Equity 66.5% below balance on 1 November | Always request equity-based drawdown and time under water |
| Account configuration matters | -4.22% to +2.75% in the same month across accounts | Ask which configuration produced the published results |
| Costs are small but real | Swap was 1.6% of net profit on account 26; -13,665.86 on account 1 in October | Include financing in every performance comparison |
9. How to evaluate any AI trading system: a checklist
Whether you are looking at PMTS or at a competitor, the same questions separate a credible report from a promotional one. We apply them to ourselves in this article, and we invite readers to apply them to us.
| Question | Why it matters | What a good answer looks like |
|---|---|---|
| Is the data pulled from a broker account or typed in by hand? | Hand-entered figures can be edited | Direct MT5 sync with a verifiable account number |
| Are losing months and losing accounts shown? | Survivorship bias hides risk | Every account and month listed, as in section 4 |
| Is equity drawdown reported alongside closed-trade drawdown? | Open positions carry risk | Both figures and the time spent under water |
| What is the payoff ratio? | Determines the break-even win rate | Disclosed, with the loss distribution |
| How many trades support the claim? | Small samples are noise | At least 100 trades, ideally several hundred |
| Are costs included? | Swap, commission and slippage reduce returns | Net results after all costs |
| Are limitations stated? | Every dataset has gaps | A section like section 2 of this article |
10. What we are watching in November
We do not forecast gold prices and we do not make promises about future returns. We can, however, say what we will be reading in the data over the coming weeks, because these are the questions that determine whether the picture in this article improves or worsens.
The first is the account 1 equity gap: whether it narrows as open positions close, widens, or stays flat. The second is the trade count on the large account, since six trades a month is too few to separate skill from luck, and a larger sample would make the monthly results far more informative. The third is account 3, which has now produced two consecutive losing months on a small number of trades and merits a configuration review. The fourth is data quality: the daily and monthly tables should update on a consistent schedule, the open-positions feed should reflect current positions, and the equity field should be consistent across accounts. These are infrastructure issues, but they decide whether outside readers can trust the dashboard.
11. Frequently asked questions
Is a 25.32% return since July 2025 a good result?
It is a reasonable result for a 15-month history with a 2% closed-trade drawdown, but it comes from a single 50,000 account and 137 trades. It is evidence, not proof. The more important number is the 55.7% break-even win rate against an 84.7% actual win rate, which shows how much room for error the strategy has had.
Why publish a month that lost money?
Because a report that only shows profitable months teaches the reader nothing about risk. Our September guide already covered the accounts that lost money, and consistency matters more than a flattering month.
Does PMTS manage client money?
PMTS is an algorithmic trading platform that connects to MetaTrader 5 accounts at brokers. The tracked accounts in this article are accounts connected to the platform. Nothing here is an offer of investment management or a solicitation to invest.
Why is the Sharpe ratio not discussed more?
Because the value in the database (7.42) is too high to be credible as an annualized figure. We prefer profit factor, drawdown and payoff structure until the calculation is audited.
Can I test the algorithm myself?
Yes. You can try the PMTS algorithm and watch the signals and the dashboard in real time, then compare the live statistics with the ones in this article.
12. Conclusion
The honest summary of October 2026 is mixed. The reference account continues to compound slowly, with a profit factor above four and a closed-trade drawdown under 2%. One of the larger accounts lost 0.74% on a single bad day. A small account lost money for a second month. And the largest account shows an equity gap that closed-trade statistics cannot explain.
A trading platform earns trust the way an auditor does: by showing the parts that do not look good alongside the parts that do. If you evaluate algorithmic trading systems for a living, use the checklist in section 9 on us. If you want to see the live dashboard behind these numbers, start a PMTS trial and compare the statistics yourself.
Risk disclosure: Trading leveraged instruments such as gold (XAUUSD) involves a high risk of loss and is not suitable for every investor. Past performance is not indicative of future results. The figures in this article are taken from the PMTS statistics database as of 1 November 2026, are shown without adjustment except where stated, and may contain reporting limitations described in section 2. Nothing in this article is investment advice.
Ready to start trading with AI?
Join hundreds of traders using PMTS algorithmic trading technology
Get Started