Profile Analysis - PolySimulator

Overview

The Profile Analysis endpoint aggregates user profile + portfolio data into a single payload for bots, dashboards, and LLM workflows. This page documents the exact calculation logic used by the API so users can verify every metric.

This endpoint is designed for MCP (Model Context Protocol) tools and AI agents. Values are rounded for display in the API response, but formulas below describe the source calculations.

Endpoint

GET

/v1/account/profile-analysis

Full profile analysis with all metrics

Query Parameters

recent_trades

integer

default:"20"

Number of recent trades to include (1-100)

equity_days

integer

default:"90"

Days of equity history to analyze (1-365)

wallet_id

integer | 'all' | 'api'

default:"api"

Wallet scope for the whole analysis — positions, trading stats, risk metrics, snapshots, recent trades, cash basis and PnL baseline alike. An integer scopes to a single wallet you own (404 WALLET_NOT_FOUND otherwise); api scopes to your API wallet (including legacy rows recorded before per-wallet attribution); all restores the cross-wallet blended view. Keywords are case-insensitive; any other value returns 422 VALIDATION_FAILED. Omitted = api.

Default changed on 2026-06-10. Before 2026-06-10 this endpoint always blended API-wallet cash with all wallets’ position values, so api_pnl could report large gains on a flat API wallet. The analysis is now scoped to the API wallet by default and api_pnl reconciles with Balance. Pass wallet_id=all if you depended on the old blended view.

Authentication

API key only — X-API-Key: <key> (or the PM-compat POLY_API_KEY / Authorization: Bearer ps_live_... aliases). A Supabase Bearer JWT is not accepted here.

curl -H "X-API-Key: $POLYSIM_API_KEY" \
  https://api.polysimulator.com/v1/account/profile-analysis

Errors

All errors return {"error": "<CODE>", "message": "<human-readable>"}.

Status	`error` code	When
401	`MISSING_API_KEY`	No API key header supplied
401	`INVALID_KEY`	API key is unknown, deactivated, or expired
401	`USER_NOT_FOUND`	Authenticated user no longer exists in the database — re-authenticate
404	`ACCOUNT_NOT_FOUND`	Authenticated user has no account record
404	`WALLET_NOT_FOUND`	Integer `wallet_id` does not exist or is not owned by the caller
422	`VALIDATION_FAILED`	`wallet_id` is neither an integer, `all`, nor `api`
500	`ANALYSIS_ERROR`	Unexpected server error building the analysis

Response

The top level also carries analysis_generated_at (ISO-8601 UTC string), analysis_version (string, currently "1.0.0"), and a natural_language_summary string, alongside these sections:

`profile` — User Metadata

Field	Type	Description
`user_id`	integer	Internal user ID
`username`	string \| null	Public username
`display_name`	string \| null	Display name
`bio`	string \| null	User bio
`avatar_url`	string \| null	Avatar image URL
`auth_provider`	string \| null	Auth provider used to sign up (e.g. `email`, `google`)
`account_created_at`	string \| null	ISO-8601 UTC registration timestamp
`account_age_days`	integer \| null	Days since registration
`api_key_tier`	string	API key tier ceiling: `free`, `pro`, or `enterprise` (a Pro+ plan maps to the `enterprise` key tier). Defaults to `free`.
`profile_visibility`	string	Profile visibility (e.g. `public`, `private`). Defaults to `public`.

`balance` — Balance Summary

Field	Type	Description
`currency`	string	Always `USD`
`ui_balance`	string	UI (MAIN) trading balance — always the MAIN wallet, regardless of `wallet_id`
`api_balance`	string	Cash of the scoped wallet (the API wallet by default; the field name is historical)
`starting_ui_balance`	string	UI baseline = seed + top-ups + grants
`starting_api_balance`	string	PnL baseline of the scoped wallet — its actual `starting_balance` (tier-aware: Pro $10,000 / Pro+$ 25,000)
`ui_pnl`	string	UI profit/loss (`ui_balance − starting_ui_balance`)
`api_pnl`	string	Scoped-wallet profit/loss (`total_portfolio_value − starting_api_balance`)
`ui_pnl_percentage`	string	UI P&L as percentage
`api_pnl_percentage`	string	Scoped-wallet P&L as percentage
`total_portfolio_value`	string	Scoped-wallet cash + scoped-wallet positions
`unrealized_pnl`	string	Open-position mark-to-market P&L (scoped)
`realized_pnl`	string	Realized P&L from closed positions (scoped)

`trading_stats` — Trading Statistics

Field	Type	Description
`total_trades`	integer	Total filled trades
`win_rate`	string	Win rate percentage
`profit_factor`	string	Gross profit / gross loss
`total_volume`	string	Total traded volume
`avg_trade_size`	string	Average trade notional
`best_trade_pnl`	string	Best single trade P&L
`worst_trade_pnl`	string	Worst single trade P&L
`unique_markets_traded`	integer	Distinct markets
`active_trading_days`	integer	Days with at least one trade

`risk_metrics` — Risk Analysis

Field	Type	Description
`portfolio_diversity_score`	string	1 - HHI (0=concentrated, 1=diverse)
`largest_position_weight`	string	Largest position as % of portfolio
`top_3_concentration`	string	Top 3 positions as % of portfolio
`max_drawdown_7d`	string	Max running-peak drawdown over 7 days
`max_drawdown_30d`	string	Max running-peak drawdown over 30 days
`cash_percentage`	string	Cash as % of total

How Metrics Are Calculated

Balance and PnL

Total portfolio value = api_balance + total_position_value (both scoped to the selected wallet — the API wallet by default)
UI PnL = ui_balance - starting_ui_balance (UI baseline = seed + topups + grants)
API PnL = total_portfolio_value - starting_api_balance
UI PnL % = ui_pnl / starting_ui_balance * 100
API PnL % = api_pnl / starting_api_balance * 100

starting_api_balance is the scoped wallet’s actual starting_balance (tier-aware: Pro

10,000 / Pro+

25,000 for the API wallet; a specific wallet’s own baseline when wallet_id=<id>), so api_pnl here agrees with GET /v1/account/balance and GET /v1/account/portfolio for the same wallet. Before 2026-06-10 this endpoint used a fixed $10,000 baseline regardless of tier — that caveat no longer applies.

Open Position Valuation

For each open position:

Cost basis = avg_entry_price * quantity
Market value = current_price * quantity (if live price exists)
If no live price is available, cost basis is used as fallback valuation.
Unrealized PnL (position) = market_value - cost_basis
Unrealized PnL (account) = sum(open_position_market_values) - sum(open_position_cost_basis)

Realized PnL (Closed Positions)

Closed positions in storage have quantity = 0, so realized PnL is reconstructed from filled orders. For each (market_id, outcome) closed position:

buy_notional = sum(BUY order notionals)
sell_notional = sum(SELL order notionals)
buy_qty = sum(BUY order quantities)
sell_qty = sum(SELL order quantities)
remaining_qty = buy_qty - sell_qty
settlement_value = exit_price * remaining_qty
position_realized_pnl = sell_notional + settlement_value - buy_notional

Then:

realized_pnl = sum(position_realized_pnl over closed positions)

This handles partial closes and settlement-style closes consistently.

Win Rate and Win/Loss Counts

Win/loss is based on realized PnL sign, not on price comparison alone:

Win: position_realized_pnl > 0
Loss: position_realized_pnl < 0
Break-even: position_realized_pnl == 0 (excluded from win/loss counts)

win_rate uses closed position count as denominator:

win_rate = wins / total_closed_positions * 100

Profit Factor and Trade PnL Stats

Gross profit = sum(all positive position_realized_pnl)
Gross loss = sum(abs(all negative position_realized_pnl))
profit_factor = gross_profit / gross_loss (when gross_loss > 0)
best_trade_pnl = max positive closed-position PnL
worst_trade_pnl = min negative closed-position PnL
avg_win_pnl = average of positive closed-position PnLs
avg_loss_pnl = average of negative closed-position PnLs

Drawdown (7d and 30d)

Drawdown is computed from portfolio_snapshots.total_value using a running peak:

Walk snapshots in chronological order
Track highest value seen so far (peak)
Compute drawdown each point: (value - peak) / peak
Maximum drawdown = most negative drawdown in the period

This avoids overstating drawdown by comparing early values to a later peak.

Portfolio Diversity and Concentration

For each open position:

Effective value = market_value (live price × quantity) when a live price exists, or cost_basis (entry price × quantity) as fallback when no live price is available.
Weight per position: w_i = effective_value / total_invested
HHI = sum(w_i^2)
portfolio_diversity_score = 1 - HHI
largest_position_weight = largest w_i
top_3_concentration = sum of top 3 w_i

Weights always sum to approximately 100% across all open positions.

Equity Returns

Using snapshots in the requested equity_days window:

return_7d: first-to-last return among snapshots in last 7 days
return_30d: first-to-last return among snapshots in last 30 days
return_all_time: first-to-last return across the full equity_days window

Each is:

period_return = (end_value - start_value) / start_value * 100

If there are fewer than two snapshots in a period, return is null.

Data Scope and Semantics

Wallet scope: every section — open_positions, trading_stats, risk_metrics, equity_timeline, category_exposure, recent_trades and the balance cash/baseline — covers the wallet selected by wallet_id (the API wallet when omitted; positions/orders recorded before per-wallet attribution count toward the API wallet). wallet_id=all spans every wallet you own. Only ui_balance / ui_pnl are always MAIN-wallet metrics regardless of scope.
open_positions contains only status=OPEN with quantity > 0
Trading stats and realized metrics use only filled orders
Category exposure is based on open-position values
Cash/invested percentages are relative to current total portfolio value

`equity_timeline` — Performance Timeline

Field	Type	Description
`return_7d`	string \| null	7-day return
`return_30d`	string \| null	30-day return
`return_all_time`	string \| null	All-time return
`peak_value` / `peak_date`	string \| null	Portfolio high watermark
`trough_value` / `trough_date`	string \| null	Portfolio low point
`current_value`	string \| null	Latest snapshot’s total value
`snapshots_available`	integer	Number of snapshots in the window (default `0`)
`latest_snapshot_at`	string \| null	ISO-8601 UTC time of the most recent snapshot

`category_exposure` — Category Breakdown

Array of objects showing exposure per market category:

[
  { "category": "crypto", "position_count": 3, "total_value": "500.00", "weight_percentage": "45.5%" },
  { "category": "sports", "position_count": 2, "total_value": "300.00", "weight_percentage": "27.3%" }
]

`recent_trades` — Recent Filled Orders

Array of the most recent filled orders (count controlled by the recent_trades query param, default 20). Each entry:

Field	Type	Description
`order_id`	integer	Order ID
`market_id`	string	Market `condition_id`
`market_question`	string \| null	Market question text
`side`	string	`BUY` or `SELL`
`outcome`	string	Outcome name
`price`	string	Fill price per share
`quantity`	string	Shares filled
`notional`	string	Total cash value of the fill
`filled_at`	string \| null	ISO-8601 UTC fill timestamp

See Trade History for the full per-trade semantics.

`natural_language_summary`

A pre-computed human-readable summary of the profile:

“trader123 is a PolySimulator paper trader who joined 45 days ago. Current API balance: $9,750.00 (started at$ 10,000.00, P&L: - $250.00, -2.50%). Has made 127 trades across 34 markets over 22 active trading days. Win rate: 62.5%. Currently holding 5 open positions worth ~$ 1,100.00. Top categories: crypto (45.5%), sports (27.3%), politics (18.2%).”

MCP Server

A standalone MCP server for profile analysis exists and will be published alongside the official SDK (coming soon). Beta-cohort members can request early access via their invite contact.

Available MCP Tools

Tool	Description
`get_profile_analysis`	Full analysis (primary tool)
`get_balance`	Quick balance check
`get_positions`	Position list with live prices
`get_trade_history`	Trade history with pagination
`get_equity_curve`	Equity curve snapshots

​Overview

​Endpoint

​Query Parameters

​Authentication

​Errors

​Response

​profile — User Metadata

​balance — Balance Summary

​trading_stats — Trading Statistics

​risk_metrics — Risk Analysis

​How Metrics Are Calculated

​Balance and PnL

​Open Position Valuation

​Realized PnL (Closed Positions)

​Win Rate and Win/Loss Counts

​Profit Factor and Trade PnL Stats

​Drawdown (7d and 30d)

​Portfolio Diversity and Concentration

​Equity Returns

​Data Scope and Semantics

​equity_timeline — Performance Timeline

​category_exposure — Category Breakdown

​recent_trades — Recent Filled Orders

​natural_language_summary

​MCP Server

​Available MCP Tools

Overview

Endpoint

Query Parameters

Authentication

Errors

Response

`profile` — User Metadata

`balance` — Balance Summary

`trading_stats` — Trading Statistics

`risk_metrics` — Risk Analysis

How Metrics Are Calculated

Balance and PnL

Open Position Valuation

Realized PnL (Closed Positions)

Win Rate and Win/Loss Counts

Profit Factor and Trade PnL Stats

Drawdown (7d and 30d)

Portfolio Diversity and Concentration

Equity Returns

Data Scope and Semantics

`equity_timeline` — Performance Timeline

`category_exposure` — Category Breakdown

`recent_trades` — Recent Filled Orders

`natural_language_summary`

MCP Server

Available MCP Tools