Data room / Successor concepts and their reviews
Version 2.1 funding model source
concepts/line_funding_model.py Open the file as written
#!/usr/bin/env python3
"""Equity Flow LINE — funding mechanics model.
Question 1: what consumer rate does the capital stack require, as a function
of utilization (u), liquidity-reserve ratio on undrawn capacity (rho), and
funding regime (pre- vs post-rated-takeout)?
Question 2: at the resulting rates, what do the three consumer archetypes
actually pay vs their alternatives (altruism check)?
Question 3: how cheap must origination be for small-draw accounts to break
even (the fee-free requirement)?
Assumptions (stated, per house diligence discipline):
- Funding regimes (blended cost of funding DRAWN balances):
pre-takeout : 60% warehouse @7.5% + 40% equity @12% -> 9.30%
post-takeout: 85% ABS @6.5% + 15% equity @12% -> 7.33%
(targets from the HomeDraw reviews: >=85% advance / <=6.5% coupon)
- Undrawn capacity: fund holds liquid reserves = rho x undrawn, earning 4.5%
cash yield vs a 10% blended hurdle -> drag spread 5.5% on reserved cash.
rho is a DESIGN CHOICE: 25% = hard contractual commitment (HELOC-like);
10% = guaranteed-core + best-efforts shell; 5% = approval-not-commitment.
- Servicing 60bps on drawn; expected loss 50bps on drawn (non-recourse tail,
CLTV-boxed); platform margin 50bps; origination cost amortized over
(avg line x u x 8yr life), avg line $150K.
- Consumer archetypes priced at the resulting rates, monthly compounding,
2% fee ELIMINATED (rate-only pricing is the altruism constraint).
"""
CASH_Y, HURDLE = 0.045, 0.10
SERV, EL, MARGIN = 0.006, 0.005, 0.005
AVG_LINE, LIFE = 150_000, 8
F_PRE = 0.60*0.075 + 0.40*0.12 # 9.30%
F_POST = 0.85*0.065 + 0.15*0.12 # 7.33%
def required_rate(u, rho, f, orig_cost=1500):
reserve_drag = rho*(1-u)/u * (HURDLE-CASH_Y)
orig_bps = orig_cost/(AVG_LINE*u*LIFE)
return f + SERV + EL + MARGIN + reserve_drag + orig_bps
print("REQUIRED CONSUMER RATE (rows: utilization u; cols: reserve ratio rho)")
for regime, f in [("PRE-TAKEOUT (yrs 0-3)", F_PRE), ("POST-TAKEOUT (scale)", F_POST)]:
print(f"\n {regime} — blended funding {f:.2%}")
print(" u \\ rho | 25% (hard line) 10% (core+shell) 5% (approval)")
for u in [0.30,0.40,0.50,0.60,0.70,0.80]:
row = " ".join(f"{required_rate(u,rho,f):>15.2%}" for rho in [0.25,0.10,0.05])
print(f" {u:.0%} | {row}")
# ---- consumer archetypes ----------------------------------------------------
def bal(draws, rate, months):
"""draws: list of (month, amt). Balance at `months` with monthly comp."""
r = rate/12
return sum(a*(1+r)**(months-m) for m, a in draws)
def archetypes(rate):
out = {}
# A. unemployment bridge: $1,000/mo x 12, repaid at yr 3
d = [(m,1000) for m in range(1,13)]
out["bridge $12K, repaid yr3"] = bal(d,rate,36)-12_000
out["bridge $12K, at sale yr10"] = bal(d,rate,120)-12_000
# B. card payoff: $40K day 1, repaid yr 6 / at sale yr 12
out["cards $40K, repaid yr6"] = bal([(0,40_000)],rate,72)-40_000
out["cards $40K, at sale yr12"] = bal([(0,40_000)],rate,144)-40_000
# C. scheduled: $1,000/mo x 60, sale yr 12
d = [(m,1000) for m in range(1,61)]
out["paycheck $60K, sale yr12"] = bal(d,rate,144)-60_000
return out
print("\nCONSUMER COST AT ACHIEVABLE RATES (interest only — no fees anywhere)")
for r in [0.0975, 0.1050]:
print(f"\n rate {r:.2%}:")
for k,v in archetypes(r).items():
print(f" {k:32s} ${v:>10,.0f}")
# alternatives for context
print("\nALTERNATIVES (same cash needs):")
print(f" cards kept at 23% (6yr carry on $40K) ${40_000*(1.23**6-1)*0:,.0f}"
f" -> min-payment carry ≈ ${40_000*0.23*6:,.0f} interest if balance held")
r=0.135/12; n=60; pmt=25_000*r/(1-(1+r)**-n)
print(f" personal loan $25K @13.5%/5y ${pmt*n-25_000:,.0f} interest + ${pmt:,.0f}/mo mandatory")
print(f" HELOC 8.5% on $40K, interest-only 6y ${40_000*0.085*6:,.0f} paid monthly OUT OF POCKET (requires income qual)")
# ---- small-account breakeven ------------------------------------------------
print("\nSMALL-ACCOUNT BREAKEVEN (post-takeout net spread ≈ "
f"{(0.0975-F_POST-SERV-EL)*10000:.0f}bps x drawn-$-years must cover origination):")
spread = 0.0975-F_POST-SERV-EL
for orig in [5300, 2500, 1500, 800]:
ddy = orig/spread
print(f" origination ${orig:>5,}: needs {ddy:>9,.0f} drawn-dollar-years "
f"(e.g., ${ddy/5:,.0f} avg drawn for 5 yrs)")