fix: a zero equity reading sized off the $100k default

Tradier returns total_equity 0.0 during its overnight maintenance window --
sixteen times in the last fourteen days, all around 01:06. Three defects
compounded so that a zero could size a position off default_equity:

  account_equity = equity or config.default_equity

`or` cannot tell "not supplied" from "read as zero", so a real 0.0 became
100000. Then the position-value cap, derived from the same equity, was
itself 0 -- and the guard

  if max_position_value is not None and max_position_value > 0:

SKIPPED the cap precisely when the cap said "no room". Finally the auto
path (~96% of approval requests) had no `pos_cap <= 0` early-out, unlike
the manual path at request_entry_approval. Net effect on a ~$250 account:
risk 5% of 100k = $5,000, capped only by liquidity. The broker would have
rejected it for buying power, which is luck rather than design.

Never fired in live hours -- every zero read landed with the market closed,
and _current_equity caches the last good value -- so this is latent, not a
realised loss. The reachable path is a boot that coincides with the
maintenance window, when no good read has been cached yet.

Distinguish None from zero, refuse to size on non-positive equity, treat a
zero cap as "no room" rather than "no cap", and mirror the manual path's
early-out. TRADING_DEFAULT_EQUITY is set to 300 on the live stack so even
the untested fallback cannot oversize.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01A2LepdJfPCfK5Q8uznjFBM

tests/unit/test_position_sizing.py

@@ -68,3 +68,46 @@ def test_rank_risk_factor_concentrates():
     assert rank_risk_factor(2, 2) == 0.5     # rank-2 half
     assert rank_risk_factor(3, 2) == 0.0     # outside top-2 -> skip
     assert rank_risk_factor(0, 2) == 1.0     # unranked -> not concentrated
+
+
+# --- zero-equity guards (2026-09-13) -----------------------------------------
+# Tradier returned total_equity 0.0 sixteen times in fourteen days, all inside
+# its ~01:06 maintenance window. `equity or default` read that real zero as
+# "not supplied" and sized off default_equity (100k) — on a $241 account that
+# is a position risking 5% of 100k. Two independent holes made it reachable:
+# the falsy-zero fallback, and a `max_position_value > 0` test that SKIPPED the
+# cap exactly when the cap was zero.
+
+def test_zero_equity_refuses_to_size():
+    """A real 0.0 reading must size nothing — never fall back to default_equity."""
+    r = calculate_position_size(
+        4.0, 3.80, equity=0.0,
+        config=PositionSizingConfig(risk_per_trade_pct=0.05, default_equity=100000),
+    )
+    assert r.position_size == 0
+    assert r.dollar_amount == 0.0
+
+
+def test_negative_equity_refuses_to_size():
+    r = calculate_position_size(4.0, 3.80, equity=-5.0, config=CFG)
+    assert r.position_size == 0
+
+
+def test_none_equity_still_uses_the_default():
+    """The documented fallback must survive: not-passed is not the same as zero."""
+    r = calculate_position_size(
+        4.0, 3.80, equity=None,
+        config=PositionSizingConfig(risk_per_trade_pct=0.01, default_equity=100000),
+    )
+    assert r.position_size > 0
+
+
+def test_zero_position_cap_blocks_the_trade():
+    """A cap of 0 means no room (no settled cash / gross budget spent), not 'no cap'."""
+    r = calculate_position_size(4.0, 3.80, equity=300, config=CFG, max_position_value=0.0)
+    assert r.position_size == 0
+
+
+def test_positive_cap_still_caps():
+    r = calculate_position_size(4.0, 3.80, equity=300, config=CFG, max_position_value=40.0)
+    assert r.position_size == 10   # 40 / 4.00