# ERN 75/25 -- Karsten Jeske, the safe-withdrawal-rate benchmark
# Idea: the portfolio is the boring part. The Early Retirement Now
#   safe-withdrawal-rate series (60+ installments, the most quantitatively
#   serious body of work in the FIRE literature) reaches a deliberately
#   unglamorous conclusion: a plain 75% US large-cap / 25% intermediate
#   Treasuries mix, rebalanced monthly, is hard to beat on the metric that
#   matters to a retiree, and the real work is done by the WITHDRAWAL RULE,
#   not by adding sleeves. This file is that benchmark, and the useful
#   counterweight to the trend/gold/all-weather builds in this directory.
# Inspired by: Karsten Jeske, Early Retirement Now, "Safe Withdrawal Rate Series".
#   https://earlyretirementnow.com/safe-withdrawal-rate-series/
# Backtest via long US history (SIM); UCITS buy-route in each line comment.
#meta capital:1000000
#meta withdraw:27000/year
#meta rebalance:30
#meta sim:on

75    SP500           # US large cap, fee-free index (buy VUAA IE00BFMXXD54 or CSPX IE00B5BMR087; PEA: SPEA)
25    IEF             # 7-10y Treasuries (UCITS: IE00B3VWN518 Acc / IDTM Dist; a euro-liability investor usually takes euro govt IEGA IE00B4WXJJ64 instead)

# ## The withdrawal rule IS the strategy
#
#   The 27,000 EUR on a million above is not a typo and not the 4% rule.
#   The series' central result is that 4% is calibrated on 30-year US
#   retirements and is too aggressive for a 50-year one, and its answer is a
#   valuation-linked rule rather than a fixed percentage: spend
#   a + b/CAPE, with a around 1.5-1.75% and b around 0.5. At the bundled
#   CAPE (about 42 in mid-2026) that gives roughly 2.7-2.9%, hence the
#   27,000. Two consequences worth internalizing before comparing this file
#   with anything else. Spending falls when valuations are rich and rises
#   when they are cheap, so the plan flexes instead of failing. And a
#   backtest at a RIGID 4% flatters diversified low-volatility books,
#   because it charges them for the sequence risk this rule sidesteps.
#   Run it both ways: this file at 27,000, then at 40,000, and read what
#   changes.

# ## What this file is NOT
#
#   It is the series' published benchmark, not the author's own book. His
#   real allocation is lower on equity (around 55%) and carries two things
#   no ETF line can express: a short-dated S&P 500 option-writing overlay
#   run on portfolio margin, and private real-estate funds. He is explicit
#   that the option strategy is not a recipe he recommends to others, so it
#   stays out of this file rather than being faked with a covered-call ETF,
#   which is a different exposure.

# ## The variant his own tables prefer
#
#   Swap 20 points of the S&P sleeve into US small-cap value (55 SP500 /
#   20 ZPRV IE00BSPLC413 / 25 IEF). Over post-1926 30-year cohorts that
#   variant beats the plain 75/25 in his published tables, and his own
#   verdict on the risk-parity portfolios is that whatever worked in them
#   came from their small-value sleeve rather than from the risk balancing.
#   Copy the file and try it; it costs one line.

# ## Where this disagrees with the rest of this directory, honestly
#
#   His headline ranking (30-year cohorts, rigid withdrawal, 0% terminal
#   value) puts 75/25 at a 3.83% failsafe against 3.40% for All Weather,
#   3.35% for the Golden Butterfly and 3.51% for the Golden Ratio, all
#   measured without small-cap value. Three things are worth knowing before
#   treating that as settled.
#   WINDOW. Those all-years failsafes are set by the 1900s-1910s cohorts,
#   when gold WAS the currency (a fixed nominal price by construction), and
#   are pressed down again by 1934-1968, when gold was $35 by decree and
#   illegal for Americans to hold until 1974 while T-bills sat rate-capped.
#   A gold sleeve cannot testify in those decades. Restricted to cohorts
#   since 1926 his own table already flips two of the three: All Weather
#   3.92% and Golden Ratio 3.95% against 75/25 at 3.83%, with 4% failure
#   rates of 1.30% and 0.19% against 1.94%.
#   HORIZON. Stretched to 50 years the ranking flips back and his case is
#   strong: cash and gold drag compounds with time, while sequence-risk
#   protection is front-loaded. Retirees with a 50-year horizon should read
#   that part of the series carefully.
#   VALUATION. His most durable finding survives every window: conditional
#   on CAPE above 20, failure rates roughly double for every portfolio
#   here, diversified or not. Diversification does not hedge out
#   overvaluation, which is precisely why the withdrawal rule above exists.
#
#   Compare for yourself, same window, same rule:
#     ./pofo examples/portfolios/ern-75-25-jeske.txt examples/portfolios/golden-butterfly.txt \
#            examples/portfolios/miller-50-40-10.txt examples/portfolios/all-weather-dalio.txt
#     ./pofo -fire examples/portfolios/ern-75-25-jeske.txt
#   The honest summary is that the two camps disagree less about the data
#   than about which century to weight, and about whether a spending rule
#   or an asset sleeve should absorb the shock.