Strategies Tab

An important premise. QANTHOS is an analysis, validation and discovery tool, not a financial advisor. The metrics you see in this tab are information describing a strategy’s past behaviour: they should be interpreted with your own experience, not read as verdicts. Thresholds such as “Profit Factor above 1.5 = solid” are reference conventions common in systematic trading literature, useful as orientation but not as law: their validity depends on the market, the timeframe, the costs and the strategy.

1. From MultiCharts to QANTHOS: the TradeExporter signal

QANTHOS doesn’t connect directly to your development software’s backtest engine: it reads report files that strategies write to disk. This decoupling is a deliberate design choice — it keeps QANTHOS independent from the platform and the development session, and lets you regenerate data whenever you want without keeping anything else open.

The bridge between the two halves is the TradeExporter signal, distributed together with QANTHOS. It’s a PowerLanguage/EasyLanguage signal you apply to the strategy’s chart in MultiCharts (or in TradeStation, with the appropriate language adjustments): every closed trade writes a line to the report file, and at startup it writes a metadata block at the top describing the strategy.

1.1 Applying the signal to the chart

The flow is the classic one for a signal:

1. Open the chart the strategy runs on in your development software.
2. Add the TradeExporter signal (the current version distributed with QANTHOS) to the chart, alongside the strategy.
3. Fill in the signal’s inputs (in particular the Meta* block, described below).
4. Let the backtest run: the signal generates the report file in the export folder.

Why TradeExporter instead of the platform’s native export? Because the file TradeExporter produces is already in the format QANTHOS expects — with the metadata block at the top and the trade-by-trade detail. It’s the format that enables all downstream metric calculations with no manual steps.

1.2 The //META: tags — describing the strategy

The way you describe a strategy to QANTHOS — ticker, OOS date, class, session, timeframe — is the metadata block that TradeExporter writes at the top of the file: a handful of //META:KEY=VALUE lines. It’s a simple block to configure and has the advantage of keeping the information inside the report, where it doesn’t get lost.

It’s one of the parts worth configuring carefully, because it’s here — and only here — that you tell QANTHOS what each strategy actually is. The tags are filled in by populating the Meta* inputs in the strategy editor, once, and from that point on the metadata always travels with the report. You’re therefore free to name your strategies however you like: the file name doesn’t need to follow any convention, because everything that matters to QANTHOS lives in the tags.

Only one field can safely stay empty: MetaTicker (in that case QANTHOS uses the ticker read from the chart). The others are worth filling in — in particular MetaOOS, without which Out-of-Sample metrics remain N/A.

You configure it once. The //META: tags live in the signal’s inputs applied to the strategy’s chart: you fill them in the first time and that’s it. From then on you can regenerate the reports as many times as you like — after new data, a re-optimization, or simply to update the history — without reconfiguring anything: every export rewrites the metadata block identically at the top of the file. The only thing left to do is reload the portfolio in QANTHOS.

In addition there’s //META:CURRENCY=, which QANTHOS writes automatically based on the ticker (default USD) and which determines the report file’s suffix.

What happens if a tag is malformed. QANTHOS doesn’t stop: it discards the single malformed field (for instance an OOS date in an unrecognized format) and at the end of the import shows you a single summary listing the ignored tags and the reason. No correction is applied automatically and no date is “invented”: if a valid OOS boundary is missing, the OOS date stays empty and that strategy’s OOS metrics remain N/A until you fix it — by correcting the tag and re-exporting, or by setting the date directly from QANTHOS (see §5).

1.3 Where the files end up

In the signal’s inputs you set two distinct folders:

• ExportPath — the folder for trade files (the reports QANTHOS will read). It’s meant to be portfolio-specific: one folder per portfolio is the tidiest way to keep them.
• StatusPath — a centralized folder for the live status files used by the Monitoring tab. Not related to strategy import.

TradeExporter generates and names report files automatically: you don’t need to worry about the name. Alongside the report it may write auxiliary files (_status.txt, _bardata.txt) which QANTHOS ignores during import — they’re skipped automatically, so you don’t need to separate them.

A note on “empty” files. A long/short strategy produces one file per side. If one of the two sides hasn’t traded in the period, its file will contain only the //META: block with no trades. This isn’t an error: QANTHOS recognizes it, skips it, and flags it in a summary line at the end of loading.

2. Setting up the portfolios folder

Once the reports are generated, you need to tell QANTHOS where to look for them. You do this once, from the menu:

Settings → Portfolio → Change Portfolios Folder

Here you select the “root” folder that contains your portfolios. The logic is simple: every subfolder of this root is a portfolio, and the .txt files inside it are that portfolio’s strategies. It’s the same folder you used as ExportPath in TradeExporter (or the folder that contains all of them, if you keep one portfolio per subfolder).

After setting the path, the Portfolio dropdown at the top of the Strategies tab will show all the portfolios found, each with the number of files it contains in parentheses.

3. Selecting and loading a portfolio

The toolbar at the top of the tab gathers everything you need to bring strategies on screen.

1. Choose the portfolio from the Portfolio dropdown. As soon as you select it, the status bar at the bottom shows the portfolio’s name, file count and path: a quick confirmation you’ve pointed to the right folder.
2. Press Load Portfolio. QANTHOS reads all the portfolio’s files in parallel (it uses all available cores, so even large pools load quickly), computes each strategy’s metrics, and populates the table.

Once loading is complete, a brief summary message appears with the number of strategies loaded and — if there are any — any warnings (skipped files, malformed //META: tags, OOS dates to fix).

Besides loading the whole portfolio, the toolbar offers:

Loading also means “recording”. In the background, without slowing down the interface, QANTHOS registers the data’s fingerprint in the Data Integrity system: it’s used to recognize, on subsequent loads, whether a strategy’s history has changed. It’s transparent to you in this tab, but it’s the reason why the first load of a new portfolio creates a “baseline”.

4. The strategies table

Once loading is done, you have in front of you the complete list of the portfolio’s strategies, one per row, with their metrics. It’s the heart of the tab, and it’s designed for two ways of reading: a quick scan for those who already know what they’re looking for, and a rich overview for those who want to understand every strategy in depth.

4.1 Identity information

The first columns describe what the strategy is, before how it performs:

• Strategy Name — the full name, from the report.
• Ticker — the market it trades (ES, NQ, GC, …).
• Class — the classification (Trend Following, Reversal, Bias), taken from the //META:CLASS tag or from the file name.
• TF — the trading session: ON for overnight, ID for intraday. (Careful not to confuse it with the TimeFrame column, which instead indicates the chart’s timeframe, e.g. 15min, 60min.)
• TimeFrame / TF2 — the main timeframe and any secondary timeframes.
• OOS Date — the date separating the In-Sample period from the Out-of-Sample one.

Why the ID/ON classification really matters. The session — ID (intraday) or ON (overnight) — isn’t just a descriptive label: QANTHOS uses it to correctly compute the margins required by the portfolio. A strategy that closes positions within the day commits the intraday margin (lower), while one that carries the position overnight commits the overnight margin (higher, and typically much more substantial). Distinguishing the two is what lets QANTHOS sum the right margins and precisely estimate the capital needed to trade the portfolio: a misclassified strategy skews that calculation. Each market’s intraday and overnight margin values live in the Tools tab (ID Margin / ON Margin columns).

4.2 Performance metrics

After that come the metrics QANTHOS computes over the entire history: profitability (Net Profit, Profit Factor, Avg Trade, Win Rate…), risk (Max Drawdown), risk-adjusted return (Sharpe Ratio, Sortino Ratio, Calmar Ratio), equity curve quality (Continuity Index, UPI) and more. The list is broad and customizable: nobody keeps every column on screen at once, and that’s exactly why Configure Columns exists (§6).

4.3 In-Sample vs Out-of-Sample metrics

This is one of the tab’s most important aspects. For QANTHOS the question isn’t just “did this strategy make money?”, but “did it keep making money after development ended?”. To answer, every relevant metric is computed twice: over the entire history and over the Out-of-Sample period alone (trades after the OOS Date).

The dedicated columns — OOS Trades, OOS Profit, OOS UPI, OOS Worst Day/Month — are the robustness thermometer: a strategy that’s excellent in-sample but flat out-of-sample is a classic overfitting signal. These columns only make sense, though, if each strategy’s OOS Date is set correctly: without a valid boundary, OOS metrics remain N/A (see §5).

Sorting the table. A click on a column header sorts strategies by that value; a second click reverses the order. It’s the fastest way to bring, for example, the strategies with the highest OOS UPI or the lowest Max Drawdown to the top.

5. Working on individual strategies

The table isn’t just for looking at: from here you inspect, correct and maintain each strategy.

5.1 Enabling and disabling

The first column is a checkbox. Checked strategies are enabled and propagate to the other modules (Portfolio Analyzer, Volatility Analyzer, Data Integrity, …); unchecked ones stay in the list but drop out of the analysis. It’s a non-destructive way to temporarily exclude a strategy without deleting it. The All / None toolbar buttons do the same thing in bulk.

5.2 Strategy detail (double-click)

A double-click on a strategy opens a detail dialog dedicated to its performance. It shows the main metrics at the top and, below, the strategy’s equity line with its drawdown curve and High Water Mark. What makes this dialog particularly useful is a vertical marker that visually splits the In-Sample data from the Out-of-Sample data: at a glance you see whether the equity kept climbing after the OOS boundary or whether it changed pace — exactly the information the numeric columns summarize, but made immediate by the chart.

5.3 Context menu (right-click)

A right-click on one or more selected strategies opens an actions menu:

Changes persist. When you correct an OOS Date or a Class from the tab, QANTHOS saves the change as a portfolio override: you’ll find it again on the next load without having to redo it. The cleanest approach, though, is still setting the //META: tags upstream, so the information travels with the report.

6. Configuring columns

The available metrics are many — dozens — and showing all of them would make the table unreadable. The Configure Columns button opens a window where you choose which columns to show and which to hide, checking the entries you’re interested in. Columns are grouped by type (identity, profitability, drawdown and risk, risk-adjusted return, continuity, worst periods, OOS metrics, performance) so you can find them easily.

The selection is per portfolio: each portfolio remembers its own column set, because an intraday pool and an overnight one are often looked at with different eyes. You can also drag headers to reorder columns to your liking: the order is remembered too.

A common approach is to keep a few “working” columns on screen (for example Net Profit, OOS Profit, OOS UPI, Max DD and whichever metric you’re currently reasoning about) and turn on the others only when needed. It’s a matter of personal convenience, not a rule: the point is reducing visual noise so you can focus on what you’re evaluating.

7. Common issues and how to avoid them

7.1 “I don’t see any portfolio in the dropdown”

This almost always means the portfolios folder isn’t set, or points to the wrong folder. Go back to Settings → Portfolio → Change Portfolios Folder and check that the root contains the portfolio subfolders with the .txt files inside.

7.2 “OOS metrics are all N/A”

It means the strategies don’t have a valid OOS Date — usually because the //META:OOS tag is empty or written in an unrecognized format. Fixes: fill in MetaOOS in TradeExporter (format YYYY-MM-DD) and re-export, or set/correct the date from the context menu (Edit OOS Date). Remember that QANTHOS never invents a boundary: it prefers to leave N/A rather than measure against an arbitrary date.

7.3 “A strategy I expected wasn’t loaded”

If the report contains only the //META: block and no trades (for example the long or short side that didn’t trade in the period), QANTHOS deliberately skips it and flags it in the end-of-load summary. This isn’t an error.

7.4 “I regenerated the reports but I still see old data”

After a re-export, reload the portfolio with Load Portfolio: QANTHOS doesn’t refresh the table automatically, so as not to consume resources while you work. You don’t need to reconfigure anything — the //META: tags are already inside the report — and if you’ve changed them, the new load will read the updated ones.

8. A typical workflow

For those starting from scratch, here’s the full sequence — meant as a reference path, not a rigid recipe:

1. In MultiCharts/TradeStation, apply the TradeExporter signal to each strategy’s chart, alongside the strategy itself.
2. Fill in the signal’s Meta* inputs (at least MetaOOS; ideally also MetaTicker, MetaClass, MetaSession, MetaTimeframe) and set ExportPath to a folder dedicated to the portfolio.
3. Run the backtests: the reports get written to the export folder.
4. In QANTHOS, set the portfolios root once from Settings → Portfolio → Change Portfolios Folder.
5. In the Strategies tab, choose the portfolio from the Portfolio menu and press Load Portfolio.
6. Read the loading summary and fix any warnings (malformed OOS dates, ignored tags).
7. Explore the table: sort by the metrics you care about, open a few strategies’ details with double-click to inspect the IS/OOS equity, and customize columns with Configure Columns.
8. Enable the strategies you want to carry into the analysis and move on to the Portfolio Analyzer, where the filters and the Builder will work on this set.

The Strategies tab does quiet but fundamental work: it turns a handful of text files scattered on disk into a coherent dataset, validated and enriched with metrics, ready for everything QANTHOS can do. The more care goes into the export phase (//META: tags filled in, correct OOS dates), the more reliable every downstream analysis will be.

9. To learn more

• Portfolio Rules — Builder, filters and portfolio construction.
• Tools — where you set per-ticker operating thresholds (e.g. Min Avg Trade) and margins.