C.001The Dynasty Calculator
How much can a family fortune spend each year and still keep up with the prices of what the rich buy? Pick a mix of stocks and bonds and a yearly draw, then pay for it by selling or by borrowing against the portfolio, through 98 years of market history.
The rich don’t live on the consumer price index. The prices of things they spend on — private school and college, luxury hotels, dining out, designer goods, sports cars, art, private flying — have risen 4.38% a year since 2000, against 2.53% for CPI (F.004). A fortune that spends a slice of itself every year has to outrun that. This calculator replays a family’s portfolio through history with its spending rising 1.8% a year faster than CPI, compounded, and asks whether the family still has as much buying power at the end as at the start.
That premium is our estimate for a basket of luxury goods and services, not a measured cost of living for rich households; if only part of a family’s budget runs at those prices, lower it under “More assumptions” (half the budget is about 0.9 points). The results are hypothetical, built from historical data: not a forecast, and not investment, tax or legal advice.
What the fortune is worth, in today’s top-of-the-K prices
Wealth after the draw, taxes and fees, divided by the family’s price level (CPI plus the luxury premium), across 2,000 simulated histories built from runs of real years since 1928. The line is the median; the bands hold the middle half and the middle 80%. Histories that run out sit on the floor.
Data
How much stock the draw needs
Chance of keeping the fortune’s buying power, by share in stocks, with everything else as set. Selling and borrowing side by side; the dots are your mix.
Data
How big the loan gets
Data
What actually happened
Data
What the calculator says
Bonds don’t keep up with the top of the K. Since 1928 the S&P 500 has returned 10.0% a year with dividends, 10-year Treasuries 4.5% and Treasury bills 3.4%. Against CPI that is 6.8%, 1.5% and 0.3% a year. Against prices rising 1.8% a year faster, it would have been 4.9% for stocks, −0.3% for bonds and −1.5% for bills, before tax. Bonds steady a portfolio, but at the top of the K they would have lost buying power on average.
In the model, a 2% draw keeps up only with a lot of stock. At the calculator’s defaults — $50M, 60% stocks, a 2% draw rising with the top of the K, federal tax at the top rates, 0.5% in fees, paid for by selling — the fortune keeps its buying power for 50 years in 35% of simulated histories and runs out in 27%. The median family ends with $26.2M in today’s top-of-the-K prices. All in stocks, it keeps up in 70% and runs out in 13%: at these settings more stock helps at every step, because the bonds’ shortfall outweighs the cushion they give in a crash. Real history was harsher. Of the 49 overlapping 50-year stretches since 1928, 5 kept their buying power at 60% stocks and 22 ran out; all in stocks, 33 kept up. Overlapping stretches share most of their years, so they add up to only a few independent histories, and every one of them includes some of the high inflation of 1966 to 1982.
The draw that keeps up three times in four is smaller than 2%. It is about 0.7% a year at 60% stocks and 1.8% all in stocks. Keeping up three times in four doesn’t rule out running out: at those draws the fortune runs out in under 1% of histories at 60% stocks and 10% all in stocks. A family that raised its spending with CPI would feel richer in the plan than in life: if its costs rise as fast as the top of the K did from 2000 to 2025, a CPI-indexed budget buys 41% of the original lifestyle after 50 years.
Borrowing instead of selling is mostly leverage. At the defaults, borrowing raises the chance of keeping up from 35% to 49%, but the fortune runs out about as often (27% of histories either way). Little of the gain is tax deferral: with no capital-gains tax on either side, borrowing still wins, 55% to 45%. The rest is the loan’s cost. The T-bill rate plus 1.25 points has averaged about 4.7% a year since 1928, less than the portfolio earned, so borrowing to spend has in effect geared the family into its own portfolio. At today’s rate, 5.20%, held for good, borrowing keeps up in 44% of histories and runs out in 29%. The loan grows with every draw and its interest and reaches the family’s 30% limit in 82% of histories. In the ones that run out, the draws and the interest use up the portfolio, and the end comes as a margin call once the loan has outgrown what is left. Interest on money borrowed to spend isn’t deductible. Under current law, gains that are never sold escape income tax at death, when the cost basis steps up; that doesn’t apply to assets in irrevocable trusts kept outside the estate, assets that do step up face estate tax above $15M a person, and the loan still has to be repaid.
Taxes, fees and the luxury premium each matter about as much. With no tax and no fees, the defaults keep their buying power in 77% of histories; with spending that rises only with CPI, in 77%; with both, in 97%. Municipal bonds, lower fees and a higher cost basis all help; the controls under “More assumptions” show how much.
What a portfolio loan costs
Securities-based lines of credit float at a spread over a short-term rate, usually SOFR (3.95% on Oct 30, 2026), and the spread falls as the line or the household gets bigger. The Fed held its target range at 3.75-4.00% on Oct 28, 2026; the prime rate is 7.00%. Published schedules, in percentage points over the benchmark:
| Lender | Benchmark | Spread by size | As of |
|---|---|---|---|
| Schwab Pledged Asset Line | SOFR | +4.40 ($100K line) to +2.40 ($2.5M+ line); up to 1.00 off for households with $10M+ at Schwab, so +1.40 | Oct 26, 2026 |
| Fidelity (U.S. Bank, Leader Bank) | SOFR | +3.10 ($100K line) to +1.90 ($3M+ line) | undated, read Nov 2, 2026 |
| Wells Fargo Priority Credit Line | SOFR | +3.20 (household assets under $2.5M) to +1.95 ($10-20M) and +1.75 ($20M+) | Nov 2023 |
| PNC Private Bank | SOFR | +1.75 (line under $1M) to +1.25 ($3-20M) | undated, read Nov 2, 2026 |
| Raymond James Bank | 1-month Term SOFR plus an adjustment | +5.25 ($100K line) to +2.25 ($10-25M) and +1.75 ($25M+); 0.25 off with $5M+ managed | Dec 2025 |
| Merrill Loan Management Account | 1-month BSBY (since discontinued) | +5.00 ($100K relationship) to +1.75 ($10M), +1.25 ($25M), +1.00 ($50M) and +0.75 ($100M+) | Mar 2024 |
| Interactive Brokers (margin loan) | Effective fed funds rate | +1.50 (to $100K) to +0.75 ($1-50M) and +0.50 ($50M+) | undated, read Nov 2, 2026 |
| Morgan Stanley Liquidity Access Line | not published | variable spreads not published; a promotional fixed rate of 5.40% for one year on new advances, through Nov 6, 2026 | Sep 21, 2026 |
J.P. Morgan, Goldman Sachs, UBS and Citi publish no spreads. Today's published schedules for the largest lines and households run SOFR + 1.25 to 1.95 points (PNC, Schwab with its discount for $10M-plus households, Fidelity, Wells Fargo, Raymond James); Merrill's March 2024 schedule, on a benchmark since discontinued, ran 1.75 points at $10M down to 0.75 above $100M. Banks don't publish what they charge $100M-plus families. The calculator assumes SOFR plus 1.75 points for $10M, 1.50 points from $25M, 1.25 points from $50M, 1.00 points from $100M, 0.75 points from $250M, so 5.20% today at $50M; the tiers from $50M up are our assumption. Lenders advance about 65% of the value of diversified stocks and 90% of Treasuries, and can call the loan at any time. Morgan Stanley alone reported $123.8B of “securities-based lending and other” wealth-management loans on Sep 30, 2026, up 20% in a year.
How the calculator works
History. Each year from 1928 to 2025 carries the S&P 500’s total return with dividends, the total return on 10-year Treasury bonds and the return on 3-month Treasury bills (Aswath Damodaran, NYU Stern), CPI-U inflation from December to December (Bureau of Labor Statistics) and the S&P 500’s lowest close of the year (Yahoo Finance). A simulated history strings together runs of consecutive real years, 8 years long on average, wrapping from 2025 back to 1928, so each year’s returns, inflation and interest rate stay together and decades like the 1970s keep their shape (a stationary block bootstrap). Everyone sees the same 2,000 histories. A different random draw moves the results by a point or two. Longer runs keep more of history’s long booms and busts intact and give somewhat better results: all in stocks, the defaults keep up in 70% of histories with 4-year runs, 70% with the 8-year runs used here and 74% with 16-year runs. “What actually happened” replays history in order instead, with each year’s own T-bill rate from the first year. The data are American, and the US had one of the best-performing stock markets of the period.
Prices. The family’s prices rise each year with that year’s CPI and a premium of 1.8% on top, compounded: the top-of-the-K basket in F.004 rose 4.38% a year and CPI 2.53% from 2000 to 2025, and 1.0438 ÷ 1.0253 is 1.018. Leaving out any one of the basket’s nine categories gives premiums from 1.3 to 2.1 points; leaving out art, an index of auction prices for blue-chip artists, gives the lowest. The basket measures what luxury goods and services cost, not what rich households spend overall: BLS’s experimental price indexes by income find little difference between the top and the bottom. Choose CPI to measure the family against ordinary prices. The fortune keeps its buying power if, after the chosen number of years, its net worth (portfolio less any loan) divided by the family’s price level is at least what it started with.
Spending. The draw is taken at the start of each year. By default it is a share of the starting fortune, raised every year with the family’s prices; under “More assumptions” it can instead be a share of each year’s net worth, which rises and falls with markets. A fortune runs out when it can’t pay the year’s draw, or when its debts exceed its assets.
Portfolio, fees and taxes. Stocks are the S&P 500 and bonds 10-year Treasuries, rebalanced each year, with fees taken yearly (0.5% by default; published schedules work out to roughly 0.1% (Vanguard) to 0.6% (bank trust departments and independent advisers) a year on a $50M account; Cerulli expects the average fee for $10M-plus clients to be about 0.66%). Taxes are 2026’s top rates for the whole history: 23.8% on qualified dividends and long-term gains, 40.8% on interest; the state options add each state’s top rate on dividends and gains: California 13.3% (top rate, above $1M of taxable income); New York State 10.9% (top rate, above $25M of taxable income; 9.65% above about $2.2M for a married couple and 10.3% above $5M); New York City 14.776% (New York State’s 10.9% plus the city’s 3.876%); Connecticut 6.99% (top rate, above $1M of taxable income for a married couple). All four tax capital gains and dividends as ordinary income and exempt Treasury interest (31 U.S.C. 3124) and their own municipal bonds. New York’s 2021 high-income rates run through tax year 2032; New York City’s current rates are set through 2026 and a bill to extend them through 2029 had passed the state Senate as of June 2026; Connecticut proposals for a capital-gains surcharge failed in 2025 and 2026. The calculator assumes today’s rates continue. Top rates fit the largest fortunes: a $10M family has a few hundred thousand dollars of taxable income a year, below where the top rates start in 2026 ($613,700 for the 20% capital-gains rate and $768,700 for the 37% income-tax rate, for a married couple), so its taxes would be lower than shown. Dividends are taxed at the S&P 500’s 2025 dividend yield (1.34% of the index at the start of the year) rather than history’s higher yields, so most of the stock return compounds untaxed until it’s sold; taxing each year’s own yield instead would lower the defaults’ chance of keeping up from 35% to 29%, and all in stocks from 70% to 64%. Sales realize the gain above average cost; reinvested income adds to cost. Municipal bonds are modeled as Treasuries paying 70% of the coupon, untaxed (in the last week of July 2026, Bloomberg’s municipal index yielded 70.84% of its Treasury index, after one of the year’s biggest weekly moves), and in-state under the state options. Rebalancing is assumed not to realize gains, and losses aren’t carried forward.
Borrowing. The family borrows each year’s draw against the portfolio and adds the interest to the loan; interest on money borrowed to spend isn’t deductible. The rate is SOFR plus the spread for the fortune’s size, starting at today’s SOFR; after the first year it floats with the Treasury-bill rate of each simulated year (or stays at today’s, if you choose). The bank lends up to 65% of the stocks and 90% of the bonds. The calculator tests the loan at the S&P 500’s lowest close of each year, with bonds on a straight line to their year-end value: if the loan is more than the bank will lend, the family sells at the low, pays tax on the gains and brings the loan down to its own limit (30% of the portfolio by default). At each year-end it also sells enough to get back under that limit.
What it leaves out. Estate and gift taxes ($15M exempt per person in 2026) and the trusts families use to manage them; the step-up in cost basis at death, which erases the tax on gains never sold but doesn’t apply to assets in irrevocable trusts outside the estate (wealth here is counted before tax on unsold gains, for both methods); private equity, real estate, art and other assets; changes in tax law; families cutting back after bad years, except under the share-of-wealth rule; and where markets start today, with the 10-year Treasury at 4.94%. History’s averages may not repeat; take points off stock returns under “More assumptions” to see what lower returns do.
Hypothetical results. The results come from a model run on historical data. They do not reflect actual investment results and are not a guarantee of future results; they change with each setting and as the data are updated. The model considers only US stocks, Treasury or municipal bonds and a securities-based loan; investments and strategies it doesn’t consider may do as well or better. Nothing here is investment, tax or legal advice.
Sources. Returns: Damodaran, Historical Returns on Stocks, Bonds and Bills, 1928–2025. Inflation: BLS CPI-U via FRED. Daily S&P 500 closes: Yahoo Finance (^GSPC). Rates: New York Fed, SOFR, EFFR, U.S. Treasury bill rates, Federal Reserve H.15, FOMC statement. Lending: the schedules above; SEC and FINRA investor alert: Securities-Backed Lines of Credit, Schwab: The Ins and Outs of Securities-Backed Borrowing, Customers Bank: SBLOC Solutions, Raymond James Bank: Securities Backed Options, FINRA: Securities-Backed Lines of Credit Explained, Osaic Personal Credit Line: asset eligibility (cash 95%, ETFs 50-85%), Wells Fargo: A guide to WFB Priority Credit Line (60% of eligible stocks and ETFs). Taxes: 26 U.S.C. 1(h) and 1(j) (Cornell LII), IRS Topic 559: Net investment income tax, IRS IR-2025-103: 2026 inflation adjustments, including amendments from the One, Big, Beautiful Bill, 26 U.S.C. 163(h): personal interest (Cornell LII), IRS Topic 505: Interest expense, 26 U.S.C. 67(h): miscellaneous itemized deductions, 26 U.S.C. 1014: basis of property acquired from a decedent (Cornell LII), California FTB: Capital gains and losses, California FTB: 2025 tax rate schedules, 31 U.S.C. 3124: Treasury obligations exempt from state tax (Cornell LII), Rev. Rul. 2023-2, Internal Revenue Bulletin 2023-16 (no step-up for irrevocable grantor trust assets outside the estate), Kiplinger: 2026 capital-gains tax thresholds, New York Tax Law 601: rates, including 2026 brackets and tax-benefit recapture, Deloitte: New York FY2026 budget extends high-income rates through 2032 (May 16, 2025), NYC Independent Budget Office: personal income tax rates (Feb 10, 2026), New York Tax Law 612: US obligations and municipal interest, Connecticut DRS: 2026 Form CT-1040ES rate tables, Connecticut OLR: capital gains tax rates in selected states (2025-R-0065), Connecticut SB 510 (2026) capital-gains surcharge: bill status. Munis: MacKay Shields Municipal Weekly (New York Life Investments), Jul 28, 2026. Fees: Vanguard Personal Advisor Wealth Management, Schwab Wealth Advisory, Chevy Chase Trust, Form ADV Part 2A (Mar 2025), Cerulli Associates (Apr 29, 2025). Loan balances: Morgan Stanley 3Q 2026 Financial Supplement. The annual series: dynasty-annual.csv.