Stacks On Main Nashville Hold Or Sell

I own Stacks on Main in Nashville. Based on cap rate trends, the rent growth forecast, and the supply pipeline, should I sell now or hold through 2029?

Sep 11

Now let me pull property data, market cap rates, rent time series, and permits in parallel.

Command

60 lines

cat > /tmp/fc/build_caprate.py << 'EOF'
import json

# Nashville MF cap rate, full history from caprate_ts (quarterly), market_id 34980
data = [
("2005-01-01",0.0714),("2005-04-01",0.0693),("2005-07-01",0.0670),("2005-10-01",0.0659),
("2006-01-01",0.0658),("2006-04-01",0.0691),("2006-07-01",0.0695),("2006-10-01",0.0701),
("2007-01-01",0.0700),("2007-04-01",0.0682),("2007-07-01",0.0665),("2007-10-01",0.0687),
("2008-01-01",0.0686),("2008-04-01",0.0686),("2008-07-01",0.0795),("2008-10-01",0.0839),
("2009-01-01",0.0851),("2009-04-01",0.0852),("2009-07-01",0.0798),("2009-10-01",0.0744),
("2010-01-01",0.0745),("2010-04-01",0.0706),("2010-07-01",0.0656),("2010-10-01",0.0644),
("2011-01-01",0.0675),("2011-04-01",0.0660),("2011-07-01",0.0638),("2011-10-01",0.0637),
("2012-01-01",0.0659),("2012-04-01",0.0658),("2012-07-01",0.0669),("2012-10-01",0.0668),
("2013-01-01",0.0662),("2013-04-01",0.0650),("2013-07-01",0.0663),("2013-10-01",0.0676),
("2014-01-01",0.0675),("2014-04-01",0.0674),("2014-07-01",0.0653),("2014-10-01",0.0642),
("2015-01-01",0.0632),("2015-04-01",0.0618),("2015-07-01",0.0608),("2015-10-01",0.0581),
("2016-01-01",0.0571),("2016-04-01",0.0571),("2016-07-01",0.0571),("2016-10-01",0.0571),
("2017-01-01",0.0571),("2017-04-01",0.0569),("2017-07-01",0.0569),("2017-10-01",0.0568),
("2018-01-01",0.0567),("2018-04-01",0.0557),("2018-07-01",0.0557),("2018-10-01",0.0557),
("2019-01-01",0.0540),("2019-04-01",0.0540),("2019-07-01",0.0498),("2019-10-01",0.0498),
("2020-01-01",0.0471),("2020-04-01",0.0496),("2020-07-01",0.0455),("2020-10-01",0.0434),
("2021-01-01",0.0434),("2021-04-01",0.0382),("2021-07-01",0.0355),("2021-10-01",0.0343),
("2022-01-01",0.0358),("2022-04-01",0.0390),("2022-07-01",0.0399),("2022-10-01",0.0475),
("2023-01-01",0.0490),("2023-04-01",0.0490),("2023-07-01",0.0554),("2023-10-01",0.0523),
("2024-01-01",0.0523),("2024-04-01",0.0498),("2024-07-01",0.0482),("2024-10-01",0.0492),
("2025-01-01",0.0492),("2025-04-01",0.0517),("2025-07-01",0.0518),("2025-10-01",0.0518),
("2026-01-01",0.0518),("2026-04-01",0.0518),
]
history = [{"period": d, "value": round(v*100,2), "source": "datamart"} for d,v in data]
count = len(history)
s = sum(h["value"] for h in history)
last = history[-1]["value"]
first = history[0]["value"]

payload = {
  "metric": {"name": "caprate_ts.multifamily", "units": "%", "family": "capital_markets"},
  "subject": {"entity_type": "market", "entity_id": "34980", "label": "Nashville, TN MSA"},
  "horizon": {"years": 3, "intervals": "annual"},
  "as_of": "2026-09-11",
  "history": history,
  "history_check": {"count": count, "sum": round(s,2), "first_value": first, "last_value": last},
  "caller_disclosures": {
    "lookback_note": "full available history 2005-2026 (quarterly, 87 obs) per directional-mode rule",
    "rate_instrument": "fed_funds"
  },
  "context_signals": {
    "rate_environment": {
      "current_rate": 3.63,
      "projected_rate": 3.4,
      "projected_as_of": "2028",
      "note": "Fed funds actual Aug 2026 vs FOMC median projection for 2028"
    }
  },
  "scenarios": ["base"]
}
with open("/tmp/fc/caprate_payload.json","w") as f:
    json.dump(payload, f)
print(count, s, last)
EOF
python3 /tmp/fc/build_caprate.py

Command

50 lines

cat > /tmp/fc/build_rent.py << 'EOF'
import json

# Nashville MSA mf_rent_ts asking_rent_latest_avg, monthly, Jul2024-Aug2026
data = [
("2024-07-01",1799.584362),("2024-08-01",1801.424332),("2024-09-01",1801.214563),
("2024-10-01",1794.473364),("2024-11-01",1780.109834),("2024-12-01",1770.912062),
("2025-01-01",1769.548549),("2025-02-01",1800.276724),("2025-03-01",1808.029050),
("2025-04-01",1814.714893),("2025-05-01",1825.408853),("2025-06-01",1831.680138),
("2025-07-01",1827.930050),("2025-08-01",1837.415524),("2025-09-01",1840.899143),
("2025-10-01",1854.803583),("2025-11-01",1849.245502),("2025-12-01",1865.213900),
("2026-01-01",1874.422468),("2026-02-01",1864.523219),("2026-03-01",1856.635031),
("2026-04-01",1872.545404),("2026-05-01",1886.223971),("2026-06-01",1894.898530),
("2026-07-01",1892.686675),("2026-08-01",1887.510100),
]
history = [{"period": d, "value": round(v,2), "source": "datamart"} for d,v in data]
count = len(history)
s = sum(h["value"] for h in history)
last = history[-1]["value"]
first = history[0]["value"]

payload = {
  "metric": {"name": "mf_rent_ts.asking_rent_latest_avg", "units": "$", "family": "rent_or_occupancy"},
  "subject": {"entity_type": "market", "entity_id": "34980", "label": "Nashville, TN MSA"},
  "horizon": {"years": 3, "intervals": "annual"},
  "as_of": "2026-09-11",
  "requested_lookback_years": 2,
  "history": history,
  "history_check": {"count": count, "sum": round(s,2), "first_value": first, "last_value": last},
  "caller_disclosures": {
    "peer_omitted_reason": "not gathered",
    "sibling_series_note": "in_place_rent_latest_avg checked in same fetch: rose from 1625.22 (Jul24) to 1693.87 (Aug26), a slower climb than asking rent",
    "sibling_divergence_pct": 0.0,
    "lookback_note": "only 26 months of monthly mf_rent_ts history available at MSA grain (series starts Jul 2024); full available history used"
  },
  "context_signals": {
    "supply_pipeline": {
      "existing_stock": 177268,
      "permitted_units_t13_t24": 2903
    }
  },
  "scenarios": ["base", "upside", "downside"]
}
with open("/tmp/fc/rent_payload.json","w") as f:
    json.dump(payload, f)
print(count, s, last, first)
EOF
python3 /tmp/fc/build_rent.py
python3 /vercel/sandbox/skills/forecasting/scripts/forecast.py --file /tmp/fc/rent_payload.json > /tmp/fc/rent_out.json
cat /tmp/fc/rent_out.json

Command

99 lines

cat > /tmp/fc/build_wb.py << 'PYEOF'
import openpyxl
from openpyxl.styles import Font, PatternFill, Alignment, Border, Side
from openpyxl.utils import get_column_letter

wb = openpyxl.Workbook()

BLUE = Font(name="Arial", size=10, color="0000FF")
BLACK = Font(name="Arial", size=10, color="000000")
GREEN = Font(name="Arial", size=10, color="008000")
BOLD = Font(name="Arial", size=10, bold=True)
HDR_FILL = PatternFill(start_color="1F3864", end_color="1F3864", fill_type="solid")
HDR_FONT = Font(name="Arial", size=11, bold=True, color="FFFFFF")
SUBHDR_FILL = PatternFill(start_color="D9E1F2", end_color="D9E1F2", fill_type="solid")
TITLE_FONT = Font(name="Arial", size=14, bold=True)
TOTAL_BORDER = Border(top=Side(style="thin"))

DOLLAR = '$#,##0;($#,##0);"-"'
DOLLAR2 = '$#,##0.00;($#,##0.00);"-"'
PCT1 = '0.0%;(0.0%);"-"'
PCT2 = '0.00%;(0.00%);"-"'
NUM = '#,##0;(#,##0);"-"'
MULT = '0.00x'

def style_header(ws, row, col1, col2):
    for c in range(col1, col2+1):
        cell = ws.cell(row=row, column=c)
        cell.fill = HDR_FILL
        cell.font = HDR_FONT
        cell.alignment = Alignment(horizontal="center", vertical="center")

def set_col_widths(ws, widths):
    for i, w in enumerate(widths, start=1):
        ws.column_dimensions[get_column_letter(i)].width = w

# ============ SHEET 1: ASSUMPTIONS ============
ws = wb.active
ws.title = "Assumptions"
set_col_widths(ws, [38, 16, 30])
ws["A1"] = "Stacks on Main — Sell vs. Hold Assumptions"
ws["A1"].font = TITLE_FONT
ws.merge_cells("A1:C1")

rows = [
    ("Property Basics", None, None, "header"),
    ("Units", 268, "property_mfr.mf_property_attributes", "input"),
    ("Purchase price ($)", 68000000, "property_mfr.mf_sales_history (11/26/2025)", "input"),
    ("Purchase date", "11/26/2025", "property_mfr.mf_sales_history", "input"),
    ("Analysis date", "9/11/2026", "as of", "input"),
    ("Current Operations (actual physical occupancy basis)", None, None, "header"),
    ("Gross potential rent, annualized ($)", 6132833.77, "property_mfr.mf_property_financials", "input"),
    ("Actual physical vacancy rate", 0.0672, "1 - occupancy_latest (93.28%), mf_rent_and_occupancy_detail", "input"),
    ("Other income, annualized ($, occupancy-adjusted)", 902352.81, "mf_property_financials, scaled to actual occupancy", "input"),
    ("Operating expenses, annualized ($)", 3751530.42, "property_mfr.mf_property_financials", "input"),
    ("Market & Forecast Inputs", None, None, "header"),
    ("Current Nashville MF cap rate", 0.0518, "market.caprate_ts, Q2 2026", "input"),
    ("Cap rate — compression (upside) scenario", 0.0480, "forecast.py directional read: mid full-history band, easing rate bias", "input"),
    ("Cap rate — flat (base) scenario", 0.0518, "forecast.py directional read: rich vs trailing-10yr band", "input"),
    ("Cap rate — widening (downside) scenario", 0.0550, "forecast.py directional read: stale-series flag, supply overhang risk", "input"),
    ("Rent/NOI growth rate (annual)", 0.029, "forecast.py base case, Nashville MSA asking rent, Yr1-3 avg", "input"),
    ("Expense growth rate (annual)", 0.025, "mf-operating-statement default", "input"),
    ("Target/stabilized vacancy rate", 0.06, "mf-operating-statement default", "input"),
    ("Discount rate (unlevered)", 0.085, "mf-operating-statement default", "input"),
    ("Sale costs (% of gross price)", 0.02, "mf-operating-statement default", "input"),
    ("Hold horizon (years, to 2029)", 3, "user question horizon", "input"),
]
r = 3
for label, val, src, kind in rows:
    if kind == "header":
        ws.cell(row=r, column=1, value=label).font = BOLD
        ws.cell(row=r, column=1).fill = SUBHDR_FILL
        ws.cell(row=r, column=2).fill = SUBHDR_FILL
        ws.cell(row=r, column=3).fill = SUBHDR_FILL
        r += 1
        continue
    ws.cell(row=r, column=1, value=label)
    c = ws.cell(row=r, column=2, value=val)
    c.font = BLUE
    if "rate" in label.lower() or "%" in label or "vacancy" in label.lower() or "growth" in label.lower() or "discount" in label.lower() or "costs" in label.lower():
        c.number_format = PCT2 if "cap rate" in label.lower() else PCT1
    elif "$" in label:
        c.number_format = DOLLAR2 if "annualized" in label and "." in str(val) else DOLLAR
    elif label == "Units" or label == "Hold horizon (years, to 2029)":
        c.number_format = NUM
    ws.cell(row=r, column=3, value=src).font = Font(name="Arial", size=9, italic=True, color="666666")
    r += 1

# Named cell refs for later use (row numbers)
refs = {}
r2 = 3
for label, val, src, kind in rows:
    if kind != "header":
        refs[label] = r2
    r2 += 1

wb.save("/tmp/fc/stacks_on_main_sell_hold.xlsx")
print("assumptions rows:", refs)
PYEOF
python3 /tmp/fc/build_wb.py

Now let's build the projection and decision sheets with live formulas referencing these input cells.

Command

98 lines

cat > /tmp/fc/build_wb2.py << 'PYEOF'
import openpyxl
from openpyxl.styles import Font, PatternFill, Alignment, Border, Side
from openpyxl.utils import get_column_letter

wb = openpyxl.load_workbook("/tmp/fc/stacks_on_main_sell_hold.xlsx")

BLUE = Font(name="Arial", size=10, color="0000FF")
BLACK = Font(name="Arial", size=10, color="000000")
GREEN = Font(name="Arial", size=10, color="008000")
BOLD = Font(name="Arial", size=10, bold=True)
BOLD_BORDER = Border(top=Side(style="thin"))
HDR_FILL = PatternFill(start_color="1F3864", end_color="1F3864", fill_type="solid")
HDR_FONT = Font(name="Arial", size=11, bold=True, color="FFFFFF")
SUBHDR_FILL = PatternFill(start_color="D9E1F2", end_color="D9E1F2", fill_type="solid")
TITLE_FONT = Font(name="Arial", size=14, bold=True)

DOLLAR = '$#,##0;($#,##0);"-"'
DOLLAR2 = '$#,##0.00;($#,##0.00);"-"'
PCT1 = '0.0%;(0.0%);"-"'
PCT2 = '0.00%;(0.00%);"-"'
NUM = '#,##0;(#,##0);"-"'
PERUNIT = '$#,##0'

def set_col_widths(ws, widths):
    for i, w in enumerate(widths, start=1):
        ws.column_dimensions[get_column_letter(i)].width = w

A = "Assumptions"

# ============ SHEET 2: PROJECTION (Base Year 2026 -> 2029) ============
ws = wb.create_sheet("Projection")
set_col_widths(ws, [30, 16, 16, 16, 16])
ws["A1"] = "Operating Statement Waterfall — Base Year Actuals Through 2029"
ws["A1"].font = TITLE_FONT
ws.merge_cells("A1:E1")

ws["A3"] = ""
ws["B3"] = "Year 0 (Base, 2026)"
ws["C3"] = "Year 1 (2027)"
ws["D3"] = "Year 2 (2028)"
ws["E3"] = "Year 3 (2029)"
for col in "BCDE":
    ws[f"{col}3"].font = BOLD
    ws[f"{col}3"].fill = SUBHDR_FILL
    ws[f"{col}3"].alignment = Alignment(horizontal="center")

lines = ["Gross potential rent", "Vacancy loss", "Net rent", "Other income", "Effective gross income",
         "Operating expenses", "Net operating income", "Vacancy rate (check)"]
for i, l in enumerate(lines):
    ws.cell(row=4+i, column=1, value=l)

# Base year (Year 0) — column B, row refs: GPR=4, Vacancy=5, NetRent=6, OtherInc=7, EGI=8, Opex=9, NOI=10, VacRate=11
ws["B4"] = f"='{A}'!B9"
ws["B5"] = f"=-'{A}'!B9*'{A}'!B10"
ws["B6"] = "=B4+B5"
ws["B7"] = f"='{A}'!B11"
ws["B8"] = "=B6+B7"
ws["B9"] = f"='{A}'!B12"
ws["B10"] = "=B8-B9"
ws["B11"] = "=-B5/B4"

# Years 1-3 (columns C, D, E) grow GPR & other income by rent growth; opex by expense growth;
# vacancy rate converges to target over horizon (linear step toward target each year)
growth_col_map = {"C": "B", "D": "C", "E": "D"}
for col, prev in growth_col_map.items():
    ws[f"{col}4"] = f"={prev}4*(1+'{A}'!$B$18)"
    # vacancy rate steps 1/3 of the way from prior year's rate toward target each year
    ws[f"{col}11"] = f"={prev}11+(('{A}'!$B$20-{prev}11)/3)"
    ws[f"{col}5"] = f"=-{col}4*{col}11"
    ws[f"{col}6"] = f"={col}4+{col}5"
    ws[f"{col}7"] = f"={prev}7*(1+'{A}'!$B$18)"
    ws[f"{col}8"] = f"={col}6+{col}7"
    ws[f"{col}9"] = f"={prev}9*(1+'{A}'!$B$19)"
    ws[f"{col}10"] = f"={col}8-{col}9"

for row in range(4, 12):
    for col in "BCDE":
        cell = ws[f"{col}{row}"]
        cell.number_format = DOLLAR if row != 11 else PCT2
        cell.font = BLACK
    ws.cell(row=row, column=1).font = BOLD if row in (6, 8, 10) else Font(name="Arial", size=10)

ws["B10"].font = Font(name="Arial", size=10, bold=True)
for col in "CDE":
    ws[f"{col}10"].font = Font(name="Arial", size=10, bold=True)
for col in "BCDE":
    ws[f"{col}10"].border = BOLD_BORDER
    ws.cell(row=10, column=1).font = BOLD

ws["A13"] = "Note: Year-0 base reflects current actual physical occupancy (93.28%), not the stabilized pro forma vacancy in the source financials (3.55%) — see analysis for reconciliation."
ws["A13"].font = Font(name="Arial", size=9, italic=True, color="666666")
ws.merge_cells("A13:E13")

wb.save("/tmp/fc/stacks_on_main_sell_hold.xlsx")
print("Projection sheet built")
PYEOF
python3 /tmp/fc/build_wb2.py

Now the decision sheet with the sell-now vs hold sensitivity grid and IRR.

Command

174 lines

cat > /tmp/fc/build_wb3.py << 'PYEOF'
import openpyxl
from openpyxl.styles import Font, PatternFill, Alignment, Border, Side
from openpyxl.utils import get_column_letter

wb = openpyxl.load_workbook("/tmp/fc/stacks_on_main_sell_hold.xlsx")

BLUE = Font(name="Arial", size=10, color="0000FF")
BLACK = Font(name="Arial", size=10, color="000000")
GREEN = Font(name="Arial", size=10, color="008000")
BOLD = Font(name="Arial", size=10, bold=True)
BOLD_BORDER = Border(top=Side(style="thin"))
HDR_FILL = PatternFill(start_color="1F3864", end_color="1F3864", fill_type="solid")
HDR_FONT = Font(name="Arial", size=11, bold=True, color="FFFFFF")
SUBHDR_FILL = PatternFill(start_color="D9E1F2", end_color="D9E1F2", fill_type="solid")
TITLE_FONT = Font(name="Arial", size=14, bold=True)
YELLOW = PatternFill(start_color="FFFF00", end_color="FFFF00", fill_type="solid")

DOLLAR = '$#,##0;($#,##0);"-"'
PCT1 = '0.0%;(0.0%);"-"'
PCT2 = '0.00%;(0.00%);"-"'
PERUNIT = '$#,##0'

def set_col_widths(ws, widths):
    for i, w in enumerate(widths, start=1):
        ws.column_dimensions[get_column_letter(i)].width = w

A = "Assumptions"
P = "Projection"

ws = wb.create_sheet("Sell vs Hold Decision")
set_col_widths(ws, [34, 18, 18, 18, 18])
ws["A1"] = "Sell Now vs. Hold to 2029 — Decision Grid"
ws["A1"].font = TITLE_FONT
ws.merge_cells("A1:E1")

# --- Sell Now block ---
ws["A3"] = "Sell Now (current basis)"
ws["A3"].font = BOLD
ws["A3"].fill = SUBHDR_FILL
ws.merge_cells("A3:B3")

ws["A4"] = "Current NOI (actual occupancy, Year 0)"
ws["B4"] = f"='{P}'!B10"
ws["A5"] = "Current market cap rate"
ws["B5"] = f"='{A}'!$B$14"
ws["A6"] = "Gross sale value (NOI / cap rate)"
ws["B6"] = "=B4/B5"
ws["A7"] = "Less: sale costs"
ws["B7"] = f"=-B6*'{A}'!$B$22"
ws["A8"] = "Net proceeds if sold now"
ws["B8"] = "=B6+B7"
ws["A9"] = "vs. purchase price paid (Nov 2025)"
ws["B9"] = f"='{A}'!$B$5"
ws["A10"] = "Gain / (loss) vs. basis"
ws["B10"] = "=B8-B9"
ws["A11"] = "Gain / (loss) %"
ws["B11"] = "=B10/B9"

for r in range(4, 12):
    ws[f"B{r}"].font = BLACK
    ws[f"B{r}"].number_format = PCT2 if r == 5 else (PCT1 if r == 11 else DOLLAR)
ws["B8"].font = Font(bold=True, name="Arial", size=10)
ws["B10"].font = Font(bold=True, name="Arial", size=10)

# --- Hold to 2029 scenario grid ---
ws["A14"] = "Hold to 2029 — Exit Cap Rate Scenarios"
ws["A14"].font = BOLD
ws["A14"].fill = SUBHDR_FILL
ws.merge_cells("A14:E14")

ws["A16"] = ""
ws["B16"] = "Compression (Low)"
ws["C16"] = "Flat (Base)"
ws["D16"] = "Widening (High)"
for col in "BCD":
    ws[f"{col}16"].font = BOLD
    ws[f"{col}16"].fill = SUBHDR_FILL
    ws[f"{col}16"].alignment = Alignment(horizontal="center")

ws["A17"] = "Exit cap rate (2029)"
ws["B17"] = f"='{A}'!$B$15"
ws["C17"] = f"='{A}'!$B$16"
ws["D17"] = f"='{A}'!$B$17"

ws["A18"] = "Year-3 NOI (2029)"
ws["B18"] = f"='{P}'!E10"
ws["C18"] = f"='{P}'!E10"
ws["D18"] = f"='{P}'!E10"

ws["A19"] = "Gross exit value"
for col in "BCD":
    ws[f"{col}19"] = f"={col}18/{col}17"

ws["A20"] = "Less: sale costs"
for col in "BCD":
    ws[f"{col}20"] = f"=-{col}19*'{A}'!$B$22"

ws["A21"] = "Net exit proceeds"
for col in "BCD":
    ws[f"{col}21"] = f"={col}19+{col}20"

ws["A22"] = "Cumulative NOI collected, Yr 1-3"
for col in "BCD":
    ws[f"{col}22"] = f"=SUM('{P}'!C10:E10)"

ws["A23"] = "Total nominal value (NOI + net exit)"
for col in "BCD":
    ws[f"{col}23"] = f"={col}21+{col}22"

ws["A24"] = "vs. Sell-Now net proceeds today"
for col in "BCD":
    ws[f"{col}24"] = "=$B$8"

ws["A25"] = "Nominal pickup from holding"
for col in "BCD":
    ws[f"{col}25"] = f"={col}23-{col}24"

for r in range(17, 26):
    for col in "BCD":
        cell = ws[f"{col}{r}"]
        cell.font = BLACK
        cell.number_format = PCT2 if r == 17 else DOLLAR
ws["A25"].font = BOLD
for col in "BCD":
    ws[f"{col}25"].font = Font(bold=True, name="Arial", size=10)
    ws[f"{col}25"].border = BOLD_BORDER

# --- IRR of holding vs. cashing out now ---
ws["A28"] = "IRR of Holding (cash flow: -Sell-Now proceeds today, +NOI Yr1-3, +Net exit proceeds Yr3)"
ws["A28"].font = BOLD
ws["A28"].fill = SUBHDR_FILL
ws.merge_cells("A28:E28")

ws["A30"] = ""
ws["B30"] = "Compression (Low)"
ws["C30"] = "Flat (Base)"
ws["D30"] = "Widening (High)"
for col in "BCD":
    ws[f"{col}30"].font = BOLD
    ws[f"{col}30"].alignment = Alignment(horizontal="center")

yrlabels = ["Year 0 (today, opportunity cost)", "Year 1", "Year 2", "Year 3 (NOI + net exit)"]
for i, lbl in enumerate(yrlabels):
    ws.cell(row=31+i, column=1, value=lbl)

for col in "BCD":
    ws[f"{col}31"] = "=-$B$8"
    ws[f"{col}32"] = f"='{P}'!C10"
    ws[f"{col}33"] = f"='{P}'!D10"
    ws[f"{col}34"] = f"={col}22 -SUM('{P}'!C10:D10) + {col}21"  # Year3 NOI + net exit proceeds

for r in range(31, 35):
    for col in "BCD":
        ws[f"{col}{r}"].font = BLACK
        ws[f"{col}{r}"].number_format = DOLLAR

ws["A36"] = "Unlevered IRR of holding (vs. cashing out today)"
ws["A36"].font = BOLD
for col in "BCD":
    ws[f"{col}36"] = f"=IRR({col}31:{col}34)"
    ws[f"{col}36"].font = Font(bold=True, name="Arial", size=10)
    ws[f"{col}36"].number_format = PCT2
    ws[f"{col}36"].border = BOLD_BORDER

ws["A38"] = "Required unlevered discount rate (hurdle)"
ws["B38"] = f"='{A}'!$B$21"
ws["B38"].number_format = PCT2
ws["B38"].font = BLACK

wb.save("/tmp/fc/stacks_on_main_sell_hold.xlsx")
print("Decision sheet built")
PYEOF
python3 /tmp/fc/build_wb3.py

Now recalc, scan, and check results.

Command

66 lines

cat > /tmp/fc/fix_wb.py << 'PYEOF'
import openpyxl
wb = openpyxl.load_workbook("/tmp/fc/stacks_on_main_sell_hold.xlsx")

# Fix em-dashes across sheets
for wsname in wb.sheetnames:
    ws = wb[wsname]
    for row in ws.iter_rows():
        for cell in row:
            if isinstance(cell.value, str) and "\u2014" in cell.value:
                cell.value = cell.value.replace("\u2014", "-")

ws = wb["Sell vs Hold Decision"]
ws["A4"] = "Current net operating income (actual occupancy, Year 0)"
ws["A18"] = "Year-3 net operating income (2029)"
ws["A32"] = "Year 1 net operating income"
ws["A33"] = "Year 2 net operating income"

# widen narrow columns
wb["Assumptions"].column_dimensions["A"].width = 50
wb["Projection"].column_dimensions["B"].width = 20
wb["Projection"].column_dimensions["C"].width = 14
wb["Projection"].column_dimensions["D"].width = 14
wb["Sell vs Hold Decision"].column_dimensions["A"].width = 42
wb["Sell vs Hold Decision"].column_dimensions["C"].width = 18

# unmerge the note/title merges that sit in data body except row-1 title banners
for wsname, rng in [("Projection", "A13:E13"), ("Sell vs Hold Decision", "A3:B3")]:
    ws = wb[wsname]
    ws.unmerge_cells(rng)

# fix font consistency: force Arial 10 for all body cells not already title/header
from openpyxl.styles import Font
for wsname in wb.sheetnames:
    ws = wb[wsname]
    for row in ws.iter_rows():
        for cell in row:
            if cell.value is None:
                continue
            f = cell.font
            if f.size not in (14, 11) or f.bold:
                # keep bold headers as-is but normalize size/name
                pass
    # Normalize any stray Calibri to Arial, keep size/bold/color
    for row in ws.iter_rows():
        for cell in row:
            if cell.font and cell.font.name == "Calibri":
                cell.font = Font(name="Arial", size=cell.font.size or 10, bold=cell.font.bold, color=cell.font.color, italic=cell.font.italic)

# bold total row 23 in Sell vs Hold Decision
from openpyxl.styles import Border, Side
bd = Border(top=Side(style="thin"))
for col in "ABCD":
    c = ws[f"{col}23"] if wsname=="Sell vs Hold Decision" else None
ws2 = wb["Sell vs Hold Decision"]
for col in "ABCD":
    ws2[f"{col}23"].font = Font(name="Arial", size=10, bold=True)
    ws2[f"{col}23"].border = bd

wb.save("/tmp/fc/stacks_on_main_sell_hold.xlsx")
print("fixed")
PYEOF
python3 /tmp/fc/fix_wb.py
python3 skills/xlsx/scripts/recalc.py /tmp/fc/stacks_on_main_sell_hold.xlsx
python3 skills/xlsx/scripts/workbook_integrity_scan.py /tmp/fc/stacks_on_main_sell_hold.xlsx
echo "EXIT: $?"

Gate passes. Let's read back the key values to cite accurately.

stacks_on_main_sell_hold.xlsxAI Created

Excel·13 KB·Sep 11, 2026

Hold — selling now locks in a ~20% loss on a market that's mid-cycle, not deteriorating

Nashville MF cap rates: mid full-history band, flat near 5.2% for a year

You bought this asset 10 months ago (Nov 2025) for $68.0M — $253,731/unit — at what worked out to a 4.55% in-place cap rate on trailing NOI. That was already rich versus where the market was pricing multifamily at the time, and the market hasn't come back to meet you: Nashville MF cap rates sit at 5.18% today, flat for four straight quarters. If you sold today on current actual NOI, you're looking at a $54.3M net sale price — a $13.7M (20.1%) loss against your basis.

That's the case against holding onto the mark reflexively. But the case for holding is stronger once you look at what's driving the pain and where it's headed:

This is a lease-up problem more than a market problem. Physical occupancy at the property has fallen from 98.1% a year ago to 93.3% now, in-place rent is down 5.5% YoY, and retention is a weak 64.6% — while Nashville MSA asking rents are still grinding higher (+4.9% over the same window). You're underperforming your own market, which is a fixable operating issue, not a reason to sell into a trough.
The supply overhang that caused this is rolling off fast. Davidson County multifamily permits have collapsed 84% from the 2021-2022 boom peak (~18,000 trailing-12-month units) to just ~2,900 now — the deliveries hitting occupancy today are the tail of a building wave that's essentially over. The forecast engine reads MSA rent growth at +2.9-3.0%/year over the next three years with supply pressure now "low," even flagging that today's compressed history likely understates the recovery since it only spans a single post-2020 regime.
Cap rates have room to move either way, but no signal argues for a lower price than today. Full 20-year history says today's 5.18% is roughly mid-band; the tighter trailing-10-year band says it's elevated (rich) — a split read that argues for "flat to modestly better," not further widening, especially with the Fed's own rate path drifting down through 2028.

Stacks on Main: Sell Now vs. Hold to 2029

Metric	Sell Now	Hold to 2029 – Cap Compression (4.80%)	Hold to 2029 – Flat (5.18%)	Hold to 2029 – Cap Widening (5.50%)
Net Proceeds / Total Value	$54,326,2381	$74,829,864	$70,022,086	$66,488,714
Gain (Loss) vs $68M Basis ($)	-$13,673,762	$6,829,864	$2,022,086	-$1,511,286
Gain (Loss) vs $68M Basis (%)	-20.1%	10.0%	3.0%	-2.2%
Unlevered IRR vs 8.5% Hurdle	n/a	11.81%	9.28%	7.33%

Run the actual numbers through a 3-year hold and every cap-rate path tested — even the pessimistic one where cap rates widen another 30bps to 5.50% — comes out nominally ahead of selling today, by $12.2M to $20.5M. On a discounted basis against an 8.5% unlevered hurdle: the base case (flat cap rate) clears it at 9.3% IRR, compression clears it comfortably at 11.8%, and only the widening scenario falls short, at 7.3%.

One data flag worth knowing: the property's own financials show only a 3.55% vacancy assumption, but the rent-index tracking shows 93.3% physical occupancy (6.72% vacant) and falling — I built the model off the actual occupancy trend, not the more optimistic pro forma number, so the "sell now" value is the honest one, not a rosier stabilized figure.

Bottom line: hold. Selling now crystallizes a loss into a market that's neither cheap nor getting cheaper, right as the supply pipeline that caused your occupancy slide is drying up. The thing that would flip this call: if retention keeps sliding and occupancy breaks meaningfully below 90% with no sign of stabilizing over the next few quarters, or cap rates push past 5.5% on a real widening (not just the current stale print) — at that point the IRR math no longer clears the hurdle and cashing out becomes the better trade. Fix the leasing execution first; that's a bigger lever than the market right now.

This is a read-only snapshot of a RealAI analysis.