diff --git a/shared/schemas/__init__.py b/shared/schemas/__init__.py
new file mode 100644
index 0000000..b6583bf
--- /dev/null
+++ b/shared/schemas/__init__.py
@@ -0,0 +1,37 @@
+"""Pydantic v2 schemas for all service message types."""
+
+from shared.schemas.trading import (
+    AccountInfo,
+    MarketSnapshot,
+    OrderRequest,
+    OrderResult,
+    PositionInfo,
+    SentimentContext,
+    TradeExecution,
+    TradeSignal,
+)
+from shared.schemas.news import RawArticle, ScoredArticle
+from shared.schemas.learning import TradeOutcomeSchema, WeightAdjustment
+from shared.schemas.auth import LoginRequest, RegisterRequest, TokenResponse
+
+__all__ = [
+    # Trading
+    "OrderRequest",
+    "OrderResult",
+    "PositionInfo",
+    "AccountInfo",
+    "TradeSignal",
+    "TradeExecution",
+    "MarketSnapshot",
+    "SentimentContext",
+    # News
+    "RawArticle",
+    "ScoredArticle",
+    # Learning
+    "TradeOutcomeSchema",
+    "WeightAdjustment",
+    # Auth
+    "RegisterRequest",
+    "LoginRequest",
+    "TokenResponse",
+]
diff --git a/shared/schemas/__pycache__/__init__.cpython-314.pyc b/shared/schemas/__pycache__/__init__.cpython-314.pyc
new file mode 100644
index 0000000..4e710f8
Binary files /dev/null and b/shared/schemas/__pycache__/__init__.cpython-314.pyc differ
diff --git a/shared/schemas/__pycache__/auth.cpython-314.pyc b/shared/schemas/__pycache__/auth.cpython-314.pyc
new file mode 100644
index 0000000..f9d9336
Binary files /dev/null and b/shared/schemas/__pycache__/auth.cpython-314.pyc differ
diff --git a/shared/schemas/__pycache__/learning.cpython-314.pyc b/shared/schemas/__pycache__/learning.cpython-314.pyc
new file mode 100644
index 0000000..ba29268
Binary files /dev/null and b/shared/schemas/__pycache__/learning.cpython-314.pyc differ
diff --git a/shared/schemas/__pycache__/news.cpython-314.pyc b/shared/schemas/__pycache__/news.cpython-314.pyc
new file mode 100644
index 0000000..148487b
Binary files /dev/null and b/shared/schemas/__pycache__/news.cpython-314.pyc differ
diff --git a/shared/schemas/__pycache__/trading.cpython-314.pyc b/shared/schemas/__pycache__/trading.cpython-314.pyc
new file mode 100644
index 0000000..030c4e0
Binary files /dev/null and b/shared/schemas/__pycache__/trading.cpython-314.pyc differ
diff --git a/shared/schemas/auth.py b/shared/schemas/auth.py
new file mode 100644
index 0000000..2dadacd
--- /dev/null
+++ b/shared/schemas/auth.py
@@ -0,0 +1,27 @@
+"""Authentication Pydantic schemas for API request/response payloads."""
+
+from pydantic import BaseModel, Field
+
+
+class RegisterRequest(BaseModel):
+    """Sent by the dashboard to begin passkey registration."""
+
+    username: str = Field(min_length=1, max_length=100)
+    display_name: str | None = None
+
+
+class LoginRequest(BaseModel):
+    """Sent by the dashboard to begin passkey authentication."""
+
+    username: str = Field(min_length=1, max_length=100)
+
+
+class TokenResponse(BaseModel):
+    """Returned after successful authentication."""
+
+    access_token: str
+    refresh_token: str
+    token_type: str = "bearer"
+    expires_in: int = Field(
+        description="Access token lifetime in seconds", default=900
+    )
diff --git a/shared/schemas/learning.py b/shared/schemas/learning.py
new file mode 100644
index 0000000..cb8bd41
--- /dev/null
+++ b/shared/schemas/learning.py
@@ -0,0 +1,32 @@
+"""Learning domain Pydantic schemas."""
+
+from datetime import datetime
+from uuid import UUID
+
+from pydantic import BaseModel, Field
+
+
+class TradeOutcomeSchema(BaseModel):
+    """Represents the evaluated outcome of a closed trade."""
+
+    trade_id: UUID
+    hold_duration_seconds: float = Field(ge=0)
+    realized_pnl: float
+    roi_pct: float
+    was_profitable: bool
+
+    model_config = {"from_attributes": True}
+
+
+class WeightAdjustment(BaseModel):
+    """Represents a strategy weight change made by the learning engine."""
+
+    strategy_id: UUID
+    strategy_name: str
+    old_weight: float
+    new_weight: float
+    reason: str
+    reward_signal: float
+    timestamp: datetime
+
+    model_config = {"from_attributes": True}
diff --git a/shared/schemas/news.py b/shared/schemas/news.py
new file mode 100644
index 0000000..e23adb1
--- /dev/null
+++ b/shared/schemas/news.py
@@ -0,0 +1,44 @@
+"""News article Pydantic schemas for Redis Stream messages."""
+
+from datetime import datetime
+
+from pydantic import BaseModel, Field
+
+
+class RawArticle(BaseModel):
+    """Published to ``news:raw`` by the news fetcher."""
+
+    source: str
+    url: str
+    title: str
+    content: str
+    published_at: datetime | None = None
+    fetched_at: datetime
+    content_hash: str
+
+    model_config = {"from_attributes": True}
+
+
+class ScoredArticle(BaseModel):
+    """Published to ``news:scored`` by the sentiment analyzer.
+
+    Inherits all fields from RawArticle conceptually plus scoring metadata.
+    """
+
+    # Original article fields
+    source: str
+    url: str
+    title: str
+    content: str
+    published_at: datetime | None = None
+    fetched_at: datetime
+    content_hash: str
+
+    # Scoring fields
+    ticker: str
+    sentiment_score: float = Field(ge=-1.0, le=1.0)
+    confidence: float = Field(ge=0.0, le=1.0)
+    model_used: str
+    entities: list[str] = Field(default_factory=list)
+
+    model_config = {"from_attributes": True}
diff --git a/shared/schemas/trading.py b/shared/schemas/trading.py
new file mode 100644
index 0000000..0a0046b
--- /dev/null
+++ b/shared/schemas/trading.py
@@ -0,0 +1,163 @@
+"""Trading-related Pydantic schemas for Redis Streams messages and API payloads."""
+
+from datetime import datetime
+from enum import Enum
+from typing import Any
+from uuid import UUID
+
+from pydantic import BaseModel, Field
+
+
+class OrderType(str, Enum):
+    MARKET = "market"
+    LIMIT = "limit"
+    STOP = "stop"
+
+
+class OrderSide(str, Enum):
+    BUY = "BUY"
+    SELL = "SELL"
+
+
+class OrderStatus(str, Enum):
+    PENDING = "PENDING"
+    FILLED = "FILLED"
+    CANCELLED = "CANCELLED"
+    REJECTED = "REJECTED"
+
+
+class SignalDirection(str, Enum):
+    LONG = "LONG"
+    SHORT = "SHORT"
+    NEUTRAL = "NEUTRAL"
+
+
+# ---------------------------------------------------------------------------
+# API request / response schemas
+# ---------------------------------------------------------------------------
+
+
+class OrderRequest(BaseModel):
+    """Submitted by the trade executor or the API to place an order."""
+
+    ticker: str
+    side: OrderSide
+    qty: float = Field(gt=0)
+    order_type: OrderType = OrderType.MARKET
+    limit_price: float | None = None
+    stop_price: float | None = None
+
+    model_config = {"from_attributes": True}
+
+
+class OrderResult(BaseModel):
+    """Returned after order submission or status query."""
+
+    order_id: str
+    ticker: str
+    side: OrderSide
+    qty: float
+    filled_price: float | None = None
+    status: OrderStatus
+    timestamp: datetime
+
+    model_config = {"from_attributes": True}
+
+
+class PositionInfo(BaseModel):
+    """Current position state — used in API responses and portfolio views."""
+
+    ticker: str
+    qty: float
+    avg_entry: float
+    current_price: float
+    unrealized_pnl: float
+    market_value: float
+
+    model_config = {"from_attributes": True}
+
+
+class AccountInfo(BaseModel):
+    """Account-level summary from the brokerage."""
+
+    equity: float
+    cash: float
+    buying_power: float
+    portfolio_value: float
+
+    model_config = {"from_attributes": True}
+
+
+# ---------------------------------------------------------------------------
+# Redis Stream message schemas
+# ---------------------------------------------------------------------------
+
+
+class TradeSignal(BaseModel):
+    """Published to ``signals:generated`` by the signal generator."""
+
+    ticker: str
+    direction: SignalDirection
+    strength: float = Field(ge=0.0, le=1.0)
+    strategy_sources: list[str]
+    sentiment_context: dict[str, Any] | None = None
+    timestamp: datetime
+
+    model_config = {"from_attributes": True}
+
+
+class TradeExecution(BaseModel):
+    """Published to ``trades:executed`` by the trade executor."""
+
+    trade_id: UUID
+    ticker: str
+    side: OrderSide
+    qty: float
+    price: float
+    status: OrderStatus
+    signal_id: UUID | None = None
+    strategy_id: UUID | None = None
+    timestamp: datetime
+
+    model_config = {"from_attributes": True}
+
+
+class OHLCVBar(BaseModel):
+    """Single OHLCV bar."""
+
+    timestamp: datetime
+    open: float
+    high: float
+    low: float
+    close: float
+    volume: float
+
+
+class MarketSnapshot(BaseModel):
+    """Snapshot of market data for a single ticker — used by strategies."""
+
+    ticker: str
+    current_price: float
+    open: float
+    high: float
+    low: float
+    close: float
+    volume: float
+    sma_20: float | None = None
+    sma_50: float | None = None
+    rsi: float | None = None
+    bars: list[dict[str, Any]] = Field(default_factory=list)
+
+    model_config = {"from_attributes": True}
+
+
+class SentimentContext(BaseModel):
+    """Aggregated sentiment for a ticker — passed to strategies."""
+
+    ticker: str
+    avg_score: float = Field(ge=-1.0, le=1.0)
+    article_count: int = Field(ge=0)
+    recent_scores: list[float] = Field(default_factory=list)
+    avg_confidence: float = Field(ge=0.0, le=1.0)
+
+    model_config = {"from_attributes": True}
diff --git a/tests/test_schemas.py b/tests/test_schemas.py
new file mode 100644
index 0000000..d44d248
--- /dev/null
+++ b/tests/test_schemas.py
@@ -0,0 +1,586 @@
+"""Tests for Pydantic schemas — serialization round-trips and validation constraints."""
+
+import uuid
+from datetime import datetime, timezone
+
+import pytest
+from pydantic import ValidationError
+
+from shared.schemas.trading import (
+    AccountInfo,
+    MarketSnapshot,
+    OrderRequest,
+    OrderResult,
+    OrderSide,
+    OrderStatus,
+    OrderType,
+    PositionInfo,
+    SentimentContext,
+    SignalDirection,
+    TradeExecution,
+    TradeSignal,
+)
+from shared.schemas.news import RawArticle, ScoredArticle
+from shared.schemas.learning import TradeOutcomeSchema, WeightAdjustment
+from shared.schemas.auth import LoginRequest, RegisterRequest, TokenResponse
+
+
+# ---------------------------------------------------------------------------
+# Trading schemas
+# ---------------------------------------------------------------------------
+
+
+class TestOrderRequest:
+    def test_valid_market_order(self) -> None:
+        o = OrderRequest(ticker="AAPL", side=OrderSide.BUY, qty=10.0)
+        assert o.order_type == OrderType.MARKET
+        assert o.limit_price is None
+
+    def test_valid_limit_order(self) -> None:
+        o = OrderRequest(
+            ticker="TSLA",
+            side=OrderSide.SELL,
+            qty=5.0,
+            order_type=OrderType.LIMIT,
+            limit_price=250.50,
+        )
+        assert o.limit_price == 250.50
+
+    def test_qty_must_be_positive(self) -> None:
+        with pytest.raises(ValidationError):
+            OrderRequest(ticker="AAPL", side=OrderSide.BUY, qty=0)
+
+    def test_qty_must_not_be_negative(self) -> None:
+        with pytest.raises(ValidationError):
+            OrderRequest(ticker="AAPL", side=OrderSide.BUY, qty=-5)
+
+    def test_serialization_round_trip(self) -> None:
+        o = OrderRequest(
+            ticker="GOOG",
+            side=OrderSide.BUY,
+            qty=3.0,
+            order_type=OrderType.STOP,
+            stop_price=100.0,
+        )
+        data = o.model_dump()
+        restored = OrderRequest.model_validate(data)
+        assert restored == o
+
+    def test_json_round_trip(self) -> None:
+        o = OrderRequest(ticker="META", side=OrderSide.SELL, qty=1.5)
+        json_str = o.model_dump_json()
+        restored = OrderRequest.model_validate_json(json_str)
+        assert restored == o
+
+
+class TestOrderResult:
+    def test_valid_result(self) -> None:
+        now = datetime.now(timezone.utc)
+        r = OrderResult(
+            order_id="ord-123",
+            ticker="AAPL",
+            side=OrderSide.BUY,
+            qty=10.0,
+            filled_price=150.25,
+            status=OrderStatus.FILLED,
+            timestamp=now,
+        )
+        assert r.filled_price == 150.25
+        assert r.status == OrderStatus.FILLED
+
+    def test_pending_no_fill(self) -> None:
+        now = datetime.now(timezone.utc)
+        r = OrderResult(
+            order_id="ord-456",
+            ticker="TSLA",
+            side=OrderSide.SELL,
+            qty=5.0,
+            status=OrderStatus.PENDING,
+            timestamp=now,
+        )
+        assert r.filled_price is None
+
+
+class TestPositionInfo:
+    def test_valid_position(self) -> None:
+        p = PositionInfo(
+            ticker="NVDA",
+            qty=20.0,
+            avg_entry=800.0,
+            current_price=850.0,
+            unrealized_pnl=1000.0,
+            market_value=17000.0,
+        )
+        assert p.market_value == 17000.0
+
+    def test_serialization_round_trip(self) -> None:
+        p = PositionInfo(
+            ticker="AMZN",
+            qty=10.0,
+            avg_entry=180.0,
+            current_price=185.0,
+            unrealized_pnl=50.0,
+            market_value=1850.0,
+        )
+        restored = PositionInfo.model_validate(p.model_dump())
+        assert restored == p
+
+
+class TestAccountInfo:
+    def test_valid_account(self) -> None:
+        a = AccountInfo(
+            equity=100_000.0,
+            cash=25_000.0,
+            buying_power=50_000.0,
+            portfolio_value=100_000.0,
+        )
+        assert a.equity == 100_000.0
+
+
+class TestTradeSignal:
+    def test_valid_signal(self) -> None:
+        now = datetime.now(timezone.utc)
+        s = TradeSignal(
+            ticker="AAPL",
+            direction=SignalDirection.LONG,
+            strength=0.85,
+            strategy_sources=["momentum", "news_driven"],
+            timestamp=now,
+        )
+        assert s.strength == 0.85
+        assert s.sentiment_context is None
+
+    def test_strength_must_be_in_range(self) -> None:
+        now = datetime.now(timezone.utc)
+        with pytest.raises(ValidationError):
+            TradeSignal(
+                ticker="AAPL",
+                direction=SignalDirection.LONG,
+                strength=1.5,
+                strategy_sources=["momentum"],
+                timestamp=now,
+            )
+
+    def test_strength_lower_bound(self) -> None:
+        now = datetime.now(timezone.utc)
+        with pytest.raises(ValidationError):
+            TradeSignal(
+                ticker="AAPL",
+                direction=SignalDirection.SHORT,
+                strength=-0.1,
+                strategy_sources=["mean_reversion"],
+                timestamp=now,
+            )
+
+    def test_json_round_trip(self) -> None:
+        now = datetime.now(timezone.utc)
+        s = TradeSignal(
+            ticker="TSLA",
+            direction=SignalDirection.SHORT,
+            strength=0.6,
+            strategy_sources=["mean_reversion"],
+            sentiment_context={"avg_score": -0.4},
+            timestamp=now,
+        )
+        restored = TradeSignal.model_validate_json(s.model_dump_json())
+        assert restored == s
+
+
+class TestTradeExecution:
+    def test_valid_execution(self) -> None:
+        now = datetime.now(timezone.utc)
+        tid = uuid.uuid4()
+        sid = uuid.uuid4()
+        e = TradeExecution(
+            trade_id=tid,
+            ticker="AAPL",
+            side=OrderSide.BUY,
+            qty=10.0,
+            price=150.0,
+            status=OrderStatus.FILLED,
+            signal_id=sid,
+            strategy_id=uuid.uuid4(),
+            timestamp=now,
+        )
+        assert e.trade_id == tid
+        assert e.signal_id == sid
+
+    def test_optional_ids(self) -> None:
+        now = datetime.now(timezone.utc)
+        e = TradeExecution(
+            trade_id=uuid.uuid4(),
+            ticker="GOOG",
+            side=OrderSide.SELL,
+            qty=5.0,
+            price=2800.0,
+            status=OrderStatus.FILLED,
+            timestamp=now,
+        )
+        assert e.signal_id is None
+        assert e.strategy_id is None
+
+
+class TestMarketSnapshot:
+    def test_valid_snapshot(self) -> None:
+        ms = MarketSnapshot(
+            ticker="AAPL",
+            current_price=150.0,
+            open=148.0,
+            high=152.0,
+            low=147.0,
+            close=150.0,
+            volume=5_000_000.0,
+            sma_20=149.5,
+            rsi=55.0,
+        )
+        assert ms.sma_50 is None
+        assert ms.bars == []
+
+    def test_with_bars(self) -> None:
+        ms = MarketSnapshot(
+            ticker="TSLA",
+            current_price=250.0,
+            open=248.0,
+            high=255.0,
+            low=245.0,
+            close=250.0,
+            volume=10_000_000.0,
+            bars=[
+                {"timestamp": "2026-01-01T10:00:00Z", "open": 248, "close": 250}
+            ],
+        )
+        assert len(ms.bars) == 1
+
+
+class TestSentimentContext:
+    def test_valid_context(self) -> None:
+        sc = SentimentContext(
+            ticker="AAPL",
+            avg_score=0.65,
+            article_count=5,
+            recent_scores=[0.5, 0.7, 0.8],
+            avg_confidence=0.85,
+        )
+        assert sc.article_count == 5
+
+    def test_score_range_validation(self) -> None:
+        with pytest.raises(ValidationError):
+            SentimentContext(
+                ticker="AAPL",
+                avg_score=1.5,  # Out of range
+                article_count=1,
+                avg_confidence=0.5,
+            )
+
+    def test_negative_score_in_range(self) -> None:
+        sc = SentimentContext(
+            ticker="TSLA",
+            avg_score=-0.8,
+            article_count=3,
+            avg_confidence=0.9,
+        )
+        assert sc.avg_score == -0.8
+
+    def test_confidence_range_validation(self) -> None:
+        with pytest.raises(ValidationError):
+            SentimentContext(
+                ticker="AAPL",
+                avg_score=0.5,
+                article_count=1,
+                avg_confidence=1.5,  # Out of range
+            )
+
+    def test_article_count_non_negative(self) -> None:
+        with pytest.raises(ValidationError):
+            SentimentContext(
+                ticker="AAPL",
+                avg_score=0.0,
+                article_count=-1,
+                avg_confidence=0.5,
+            )
+
+
+# ---------------------------------------------------------------------------
+# News schemas
+# ---------------------------------------------------------------------------
+
+
+class TestRawArticle:
+    def test_valid_article(self) -> None:
+        now = datetime.now(timezone.utc)
+        a = RawArticle(
+            source="reuters",
+            url="https://reuters.com/article/1",
+            title="Market Rally Continues",
+            content="Stocks rose sharply today...",
+            published_at=now,
+            fetched_at=now,
+            content_hash="sha256abcdef1234567890",
+        )
+        assert a.source == "reuters"
+
+    def test_published_at_optional(self) -> None:
+        now = datetime.now(timezone.utc)
+        a = RawArticle(
+            source="reddit",
+            url="https://reddit.com/r/stocks/1",
+            title="DD on TSLA",
+            content="Here is my analysis...",
+            fetched_at=now,
+            content_hash="hash123",
+        )
+        assert a.published_at is None
+
+    def test_json_round_trip(self) -> None:
+        now = datetime.now(timezone.utc)
+        a = RawArticle(
+            source="yahoo",
+            url="https://finance.yahoo.com/1",
+            title="Earnings Beat",
+            content="Apple beat earnings...",
+            published_at=now,
+            fetched_at=now,
+            content_hash="hash456",
+        )
+        restored = RawArticle.model_validate_json(a.model_dump_json())
+        assert restored == a
+
+    def test_required_fields(self) -> None:
+        with pytest.raises(ValidationError):
+            RawArticle(source="reuters")  # type: ignore[call-arg]
+
+
+class TestScoredArticle:
+    def test_valid_scored_article(self) -> None:
+        now = datetime.now(timezone.utc)
+        sa = ScoredArticle(
+            source="reuters",
+            url="https://reuters.com/article/1",
+            title="Apple Earnings Beat",
+            content="Apple reported...",
+            published_at=now,
+            fetched_at=now,
+            content_hash="hash789",
+            ticker="AAPL",
+            sentiment_score=0.85,
+            confidence=0.92,
+            model_used="finbert",
+            entities=["Apple Inc", "Tim Cook"],
+        )
+        assert sa.sentiment_score == 0.85
+        assert sa.entities == ["Apple Inc", "Tim Cook"]
+
+    def test_sentiment_score_range(self) -> None:
+        now = datetime.now(timezone.utc)
+        with pytest.raises(ValidationError):
+            ScoredArticle(
+                source="reuters",
+                url="https://reuters.com/1",
+                title="Test",
+                content="Test content",
+                fetched_at=now,
+                content_hash="hash",
+                ticker="AAPL",
+                sentiment_score=1.5,  # Out of range
+                confidence=0.5,
+                model_used="finbert",
+            )
+
+    def test_confidence_range(self) -> None:
+        now = datetime.now(timezone.utc)
+        with pytest.raises(ValidationError):
+            ScoredArticle(
+                source="reuters",
+                url="https://reuters.com/1",
+                title="Test",
+                content="Test content",
+                fetched_at=now,
+                content_hash="hash",
+                ticker="AAPL",
+                sentiment_score=0.5,
+                confidence=-0.1,  # Out of range
+                model_used="finbert",
+            )
+
+    def test_negative_sentiment(self) -> None:
+        now = datetime.now(timezone.utc)
+        sa = ScoredArticle(
+            source="reddit",
+            url="https://reddit.com/1",
+            title="Bad news",
+            content="Terrible quarter...",
+            fetched_at=now,
+            content_hash="hashN",
+            ticker="TSLA",
+            sentiment_score=-0.9,
+            confidence=0.8,
+            model_used="ollama",
+        )
+        assert sa.sentiment_score == -0.9
+
+    def test_json_round_trip(self) -> None:
+        now = datetime.now(timezone.utc)
+        sa = ScoredArticle(
+            source="yahoo",
+            url="https://yahoo.com/1",
+            title="Headline",
+            content="Body text",
+            fetched_at=now,
+            content_hash="hashRT",
+            ticker="NVDA",
+            sentiment_score=0.3,
+            confidence=0.7,
+            model_used="finbert",
+            entities=["NVIDIA"],
+        )
+        restored = ScoredArticle.model_validate_json(sa.model_dump_json())
+        assert restored == sa
+
+
+# ---------------------------------------------------------------------------
+# Learning schemas
+# ---------------------------------------------------------------------------
+
+
+class TestTradeOutcomeSchema:
+    def test_valid_outcome(self) -> None:
+        o = TradeOutcomeSchema(
+            trade_id=uuid.uuid4(),
+            hold_duration_seconds=14400.0,
+            realized_pnl=250.50,
+            roi_pct=2.5,
+            was_profitable=True,
+        )
+        assert o.was_profitable is True
+        assert o.hold_duration_seconds == 14400.0
+
+    def test_hold_duration_non_negative(self) -> None:
+        with pytest.raises(ValidationError):
+            TradeOutcomeSchema(
+                trade_id=uuid.uuid4(),
+                hold_duration_seconds=-1.0,
+                realized_pnl=100.0,
+                roi_pct=1.0,
+                was_profitable=True,
+            )
+
+    def test_losing_trade(self) -> None:
+        o = TradeOutcomeSchema(
+            trade_id=uuid.uuid4(),
+            hold_duration_seconds=3600.0,
+            realized_pnl=-150.0,
+            roi_pct=-3.0,
+            was_profitable=False,
+        )
+        assert o.was_profitable is False
+        assert o.realized_pnl == -150.0
+
+    def test_json_round_trip(self) -> None:
+        o = TradeOutcomeSchema(
+            trade_id=uuid.uuid4(),
+            hold_duration_seconds=7200.0,
+            realized_pnl=500.0,
+            roi_pct=5.0,
+            was_profitable=True,
+        )
+        restored = TradeOutcomeSchema.model_validate_json(o.model_dump_json())
+        assert restored == o
+
+
+class TestWeightAdjustment:
+    def test_valid_adjustment(self) -> None:
+        now = datetime.now(timezone.utc)
+        wa = WeightAdjustment(
+            strategy_id=uuid.uuid4(),
+            strategy_name="momentum",
+            old_weight=0.33,
+            new_weight=0.38,
+            reason="Positive reward signal from recent trades",
+            reward_signal=0.72,
+            timestamp=now,
+        )
+        assert wa.old_weight == 0.33
+        assert wa.new_weight == 0.38
+
+    def test_required_fields(self) -> None:
+        with pytest.raises(ValidationError):
+            WeightAdjustment(
+                strategy_id=uuid.uuid4(),
+                strategy_name="momentum",
+            )  # type: ignore[call-arg]
+
+    def test_json_round_trip(self) -> None:
+        now = datetime.now(timezone.utc)
+        wa = WeightAdjustment(
+            strategy_id=uuid.uuid4(),
+            strategy_name="mean_reversion",
+            old_weight=0.30,
+            new_weight=0.25,
+            reason="Poor recent performance",
+            reward_signal=-0.4,
+            timestamp=now,
+        )
+        restored = WeightAdjustment.model_validate_json(wa.model_dump_json())
+        assert restored == wa
+
+
+# ---------------------------------------------------------------------------
+# Auth schemas
+# ---------------------------------------------------------------------------
+
+
+class TestRegisterRequest:
+    def test_valid_registration(self) -> None:
+        r = RegisterRequest(username="trader1", display_name="Top Trader")
+        assert r.username == "trader1"
+        assert r.display_name == "Top Trader"
+
+    def test_display_name_optional(self) -> None:
+        r = RegisterRequest(username="trader2")
+        assert r.display_name is None
+
+    def test_username_required(self) -> None:
+        with pytest.raises(ValidationError):
+            RegisterRequest(username="")  # min_length=1
+
+    def test_username_max_length(self) -> None:
+        with pytest.raises(ValidationError):
+            RegisterRequest(username="x" * 101)  # max_length=100
+
+
+class TestLoginRequest:
+    def test_valid_login(self) -> None:
+        l = LoginRequest(username="trader1")
+        assert l.username == "trader1"
+
+    def test_username_required(self) -> None:
+        with pytest.raises(ValidationError):
+            LoginRequest(username="")
+
+
+class TestTokenResponse:
+    def test_valid_response(self) -> None:
+        t = TokenResponse(
+            access_token="eyJ...",
+            refresh_token="eyR...",
+        )
+        assert t.token_type == "bearer"
+        assert t.expires_in == 900  # default 15 min
+
+    def test_custom_expiry(self) -> None:
+        t = TokenResponse(
+            access_token="eyJ...",
+            refresh_token="eyR...",
+            expires_in=3600,
+        )
+        assert t.expires_in == 3600
+
+    def test_json_round_trip(self) -> None:
+        t = TokenResponse(
+            access_token="access123",
+            refresh_token="refresh456",
+            token_type="bearer",
+            expires_in=1800,
+        )
+        restored = TokenResponse.model_validate_json(t.model_dump_json())
+        assert restored == t