CEF app tab: full 239-fund verified tax-arb ranking

app.py

@@ -1110,51 +1110,36 @@ with tab_fundlab:
             "table is a rough keyword classification. N-PORT holdings are "
             "quarterly and up to ~60 days stale.")
 
-    # --- closed-end funds (tax-arb shortlist) ---------------------------
+    # --- closed-end funds (tax-arb, actual distribution character) ------
     import pandas as _pd
     _fl_dir = Path(__file__).parent / "fundlab"
-    _char_path = _fl_dir / "cef_character.json"
-    if _char_path.exists():
-        _char = json.loads(_char_path.read_text())
-        _annual = {}
-        _ann_path = _fl_dir / "cef_annual.json"
-        if _ann_path.exists():
-            _annual = json.loads(_ann_path.read_text())
+    _rank_path = _fl_dir / "cef_rank_all.json"
+    if _rank_path.exists():
+        _rank = json.loads(_rank_path.read_text())
+        _univ = json.loads((_fl_dir / "cef_universe.json").read_text())
         _rows = []
-        for _s, _c in _char.items():
-            _a = _annual.get(_s.upper(), {})
-            _ch = _c.get("character")
-            _actual = False
-            if _a.get("char_actual") is not None:
-                _ch = _a["char_actual"]
-                _actual = True
-            _t5, _vol = _c.get("t5"), _c.get("vol5")
-            _score = (None if _ch is None or _t5 is None or _vol is None
-                      else round(_ch * (_t5 + 0.4 * _vol), 4))
-            _disc = _a.get("disc_now_approx")
+        for _score, _s, _ch, _t5, _vol, _npos, _tend, _disc in _rank:
             _rows.append({
-                "sym": _s.upper(),
-                "name": (_a.get("name") or _c.get("name") or "")[:44],
+                "sym": _s,
+                "name": (_univ.get(_s, {}).get("name") or "")[:44],
                 "char": _ch,
-                "actual": _actual,
                 "disc%": None if _disc is None else round(100 * _disc, 1),
                 "t5%": None if _t5 is None else round(100 * _t5, 1),
                 "vol5%": None if _vol is None else round(100 * _vol, 1),
-                "pos_scen": _c.get("n_pos_scen"),
-                "tenders": _a.get("n_tender"),
+                "pos_scen": _npos,
+                "tenders": _tend,
                 "score": _score})
         _cef_df = _pd.DataFrame(_rows)
         st.divider()
-        st.subheader("Closed-end funds — tax-arb shortlist")
-        st.dataframe(
-            _cef_df.sort_values("score", ascending=False, na_position="last"),
-            width="stretch", height=460)
+        st.subheader("Closed-end funds — tax-arb ranking")
+        st.dataframe(_cef_df, width="stretch", height=460)
         st.caption(
-            "char = distribution character 0–1 (0 = all ordinary income, 1 = "
-            "all capital-gains/ROC). actual = verified from the fund's own "
-            "per-share financial highlights (divs + gains + ROC == total "
-            "distributions per year, plus the full NAV chain); otherwise the "
-            "return-sleeve model. disc% = market vs NAV. pos_scen = severe-"
-            "drawdown scenarios with a positive 5y return. score = char × "
-            "(5y total return + 0.4 × 5y vol) — harvestable character "
-            "weighted by damage potential.")
+            f"{len(_cef_df)} CEFs with VERIFIED per-share distribution "
+            "character (divs + gains + ROC == total distributions per year, "
+            "plus the NAV chain, from the fund's own report). char 0–1: "
+            "0 = all ordinary income, 1 = all capital-gains/ROC (taxable "
+            "income that defers to the LTCG/ROC rate in a taxable account). "
+            "score = char × (5y total return + 0.4 × 5y vol): harvestable "
+            "character weighted by how much damage a forced sale could do. "
+            "tenders = N-23C repurchase/tender filings (sponsor capital-"
+            "management pressure). disc% = FY-end market vs NAV.")