White Paper · April 2026
The VistaCurve Correction Equation
A closed-form, finite-horizon correction to capital market return assumptions — validated across 720 million simulated paths and implemented across Vistamark’s quantitative platform.
Matthew Rice, CFA, CAIA — Managing Partner & Chief Investment Officer, Vistamark Investments LLC · April 2026
Imagine you’re planning a road trip. The speed limit is 60 mph — but with traffic, stop lights, and the occasional wrong turn, you’ll average something less over the full journey. The longer the drive, the more those interruptions compound against you. Now imagine every GPS in America was programmed to show 60 mph as your expected average, regardless of trip length. A one-hour drive? Fine. A ten-hour drive? You’re arriving much later than promised.
That is the glaring mathematical error in virtually every capital markets assumptions table in the industry. The “speed limit” is the arithmetic return. The profession has known for decades how to calculate the gap between that and what investors actually compound at — but only for an infinitely long trip. What it has never had is the right answer for your trip: five years, ten years, twenty years. It hides inside Monte Carlo simulations, portfolio optimizers, and return forecasts. It is not a rounding issue or an edge case — it is baked into the standard methodology. And it has been hiding in plain sight for three hundred years.
“In theory, there is no difference between theory and practice. In practice, there is.”— Yogi Berra
Yogi had a gift for landing on mathematical truth through the back door of common sense. Much of the investment profession, in practice, treats arithmetic and geometric returns as more or less interchangeable — not because anyone believes they are the same, but because the correct conversion for a finite horizon has never been cleanly solved and embedded in standard tooling. The result is a consistent overstatement of expected portfolio outcomes. Not occasionally. Every time. In every forecast that uses the wrong conversion.
The error is not subtle. It is structural. And the fix is a single equation — analytically derived, covering ground the quantitative finance literature has not, to our knowledge, previously mapped.
This paper presents that equation — the VistaCurve Correction Equation — explains the three centuries of mathematical and financial thought that made it possible, and shows what it looks like applied to a real capital market assumptions table. That last part is just one example of a much wider set of applications; it is where the abstract becomes concrete, not where the story ends. We have embedded the formula throughout VistaBuilder™ and VistaBalancer™, and its implications run deeper than any single table can capture. Pull up a chair. There will be math. There will also be baseball.
The Claim in Plain Language
This paper provides a closed-form expression for the finite-horizon geometric return G(T) that (i) matches the classical infinite-horizon lognormal result as T → ∞, and (ii) recovers the arithmetic mean exactly at T = 1 — eliminating the systematic horizon mismatch embedded in standard CMA and optimizer practice. The formula assumes lognormal return dynamics consistent with the continuous-time framework of Merton and Samuelson; within that framework, both boundary conditions hold exactly, and Monte Carlo residuals across 720 million simulated paths do not exceed 0.11%.
A Note from the Author — Twenty-Five Years in the Making
Nearly a quarter century ago — late 2002, to be precise — I was constructing my first formal 10-year Capital Market Assumptions and became genuinely obsessed with a question I couldn’t let go of: what is the actual mathematical relationship between arithmetic and geometric returns? Not the textbook answer — the real one. So I did what any self-respecting, slightly reckless young Chief Investment Officer does when he doesn’t know the answer and doesn’t know who to ask: I built a Monte Carlo simulator in Microsoft Excel. An enormous, creaking, ambitious thing that crashed my cutting-edge computer (cutting-edge for the time, anyway) dozens of times before it finally ran to completion. When it did, I understood the relationship viscerally in a way no textbook had managed to teach me.
Throughout those years, I had the great fortune of working alongside Bill Schneider — co-founder of DiMeo Schneider & Associates and, for me personally, one of the most genuinely formative mentors a young investment professional could hope for. Bill had watched me pour an almost embarrassing amount of time and energy into that Excel simulator, and when I showed him what I’d built, his reaction was perfectly calibrated: he appreciated the instinct and the effort, and then he laughed at me — warmly, but without mercy — for having spent all that time constructing a solution when a well-established answer had been sitting in the literature for decades. He walked me through the σ²/2 approximation: the classical relationship between arithmetic and geometric returns that the profession had long relied upon. It was elegant. It was right. And it was, I immediately sensed, incomplete. The formula assumed an infinite time horizon. Our clients, last I checked, are not immortal. The infinite-horizon result was a mathematical convenience dressed up as a planning tool — and Bill, to his credit, never suggested it was anything more. The gap he pointed me toward was one I’d spend the next twenty-five years trying to close.
I looked everywhere. The Journal of Finance. The Review of Financial Studies. The Financial Analysts Journal. Every major quantitative text that crossed my desk across more than two decades and well over $260+ billion in aggregate fiduciary responsibility. I never found a clean, closed-form, finite-horizon answer. The VistaCurve Correction Equation is that answer — not published in a journal, but worked out over many years of sitting with the same stubborn question: what is the right geometric return for this investor, with this horizon? I knew the answer had to involve some blend of time and volatility. I assumed the volatility component would be quadratic in nature — because variance drag compounds that way. What I wasn’t sure about was the time component. Could it really be reciprocal? That felt almost too clean. It turns out it is exactly that clean. The correction term is (1−1/T)·σ²/2: quadratic in volatility, reciprocal in horizon, interpolating perfectly between the two boundary conditions that had to be true by construction.
The formal validation came through rigorous Monte Carlo simulation — 720 million simulated lognormal return paths across 9 portfolio mixes and 8 time horizons, with formula residuals that do not exceed 0.11% under any tested parameters. More recently, the advent of AI has made independent verification available to anyone. The formal simulation work was run through Perplexity. I have since also asked Claude, Gemini, and other AI systems to run their own simulations and test the formula — and across every run that completed, the result has held up. Some problems solve themselves, eventually, if you wait long enough and the technology catches up.
Matthew Rice, CFA, CAIA | Managing Partner & CIO, Vistamark Investments LLC
Notation Used Throughout This Paper
| μA | Arithmetic mean return (simple average of periodic returns) |
| G(T) | Finite-horizon geometric return over T years — the compound annual growth rate a T-year investor actually realizes |
| G(∞) | Perpetual (infinite-horizon) geometric return — the steady-state compound rate in the limit T → ∞ |
| σ | Annualized standard deviation of returns |
| σ²/2 | Variance drag — the full penalty separating μA from G(∞) under lognormal dynamics |
| σ²/(2T) | VistaCurve correction — the horizon-specific fraction of variance drag that has not yet compounded away |
The Practical Punchline
What This Means for Your Portfolio Right Now: Vistamark’s 2026–2035 Capital Market Assumptions
Before the history. Before the derivation. Before 720 million simulated paths. Here is what the VistaCurve formula actually does to a real capital market assumptions table — applied to Vistamark’s own 2026–2035 CMAs, across twenty-three asset classes, at six investment horizons from one year to one hundred.
Vistamark’s 2026–2035 Capital Market Assumptions present ten-year geometric return forecasts for twenty-three asset classes, derived from a combination of current market data, regression-based modeling, the Black-Litterman methodology, and factor-based premiums for private markets. In our opinion, these are the best available estimates of what investors should expect from each asset class over the coming decade.
For this analysis, we treat each CMA geometric return as a perpetual long-run expectation — the steady-state rate a given asset class would deliver across all time horizons, G(∞). From this baseline, the VistaCurve formula generates horizon-specific corrections with mathematical precision. Since the CMA geometric returns represent G(∞), we solve for the implied arithmetic mean — μA = G(∞) + σ²/2 — and substitute back into the formula. A clean finite-horizon form emerges:
How to Implement — Three Steps for Any CMA Workflow
Step 1. Take the published CMA geometric return for an asset class as G(∞).
Step 2. Compute the implied arithmetic mean: μA = G(∞) + σ²/2.
Step 3. For any finite horizon T, compute G(T) = G(∞) + σ²/(2T).
No changes to underlying asset-class research are required — only the arithmetic-to-geometric conversion step is different.
Exhibit CMA — VistaCurve Horizon-Specific Geometric Return Forecasts Applied to Vistamark 2026–2035 Capital Market Assumptions · G(T) = G(∞) + σ²/(2T)
Formula: G(T) = G(∞) + σ²/(2T). CMA geometric returns treated as perpetual baselines G(∞). Standard deviations from Vistamark 2026–2035 CMAs. G(1) = μA (arithmetic return) by identity. G(10) highlighted as the most common CMA reference horizon. For illustrative purposes only. Not investment advice.
What the Numbers Tell Us
The G(1) column confirms by identity that every arithmetic return assumption in any CMA table is simply the single-year geometric return. The gap between G(1) and G(∞) — widest for Venture Capital (16.47% vs. 12.55%) and Private Equity (12.43% vs. 9.55%) — is the full variance drag that accumulates across an infinite horizon.
The 3-year column is critical for private market allocations. G(3) for Private Equity is 10.51% — nearly a full percent above G(10) of 9.84% and almost 100 basis points above the perpetual G(∞) of 9.55%. Fund-level return modeling that ignores this correction is systematically understating geometric expectations for early vintages.
Low-volatility asset classes are essentially horizon-invariant. Core bonds and most credit products show corrections under 20 basis points even at a 1-year horizon. For equities and private markets, the correction is not a rounding exercise.
The G(100) column confirms the formula’s integrity. At T = 100, corrections are negligible everywhere. Endowments modeling multi-generational portfolios in true perpetuity are well-served by the standard CMA geometric returns. Every other investor benefits from this table.
Section I
The Foundational Problem
Arithmetic vs. Geometric Returns
The arithmetic mean and the compound growth rate are not the same number. The larger the volatility and the longer the horizon, the larger the gap — and the larger the error from ignoring it.
The arithmetic mean of a set of returns is the simple average. The geometric mean is the compound annual growth rate — the return your client’s account actually compounds at. They are not the same number, and the gap between them grows with both volatility and time horizon.
The Classic Illustration. +50% then −50%: Arithmetic mean = 0%. Geometric mean = −13.4%. A client started with $1,000,000, fell to $500,000, and recovered only to $750,000 — a loss of $250,000 that the arithmetic mean reports as zero. This is not a trick. It is a mathematical identity. Losing 50% requires a 100% gain to break even. The mathematics are fundamentally asymmetric, and the greater the volatility, the larger the arithmetic-geometric gap.
Proposition 1 — The Classical Infinite-Horizon Approximation
μG(∞) ≈ μA − σ²/2
From the moment-generating function of the lognormal distribution. If ln(1+r) ~ 𝒩(α, σ²), then μA ≈ α + σ²/2 and μG(∞) ≈ α, from which the drag term σ²/2 is recovered. Note the critical assumption: T → ∞. This is the approximation the profession has used for decades. It is not wrong. It is incomplete.
Harry Markowitz (1952) built modern portfolio theory on arithmetic expected returns — a correct choice for single-period optimization that became a compounding problem when applied to multi-period forecasting. Paul Samuelson (1969) formalized the distinction, and John L. Kelly Jr. (1956) showed that maximizing the expected logarithm of wealth — equivalent to maximizing the geometric growth rate — dominates all other long-run strategies. The insights were there. The profession chose arithmetic for convenience — and the gap between theory and practice has persisted ever since.
Geometric returns reflect financial reality — the investment return your client can spend. Maximizing time-horizon geometric returns should be the objective.— Matthew Rice, CFA, CAIA | Vistamark Investments LLC
Section II
The Finite Horizon Problem
What Everyone Missed
Infinite-horizon math applied to finite-horizon clients introduces a systematic, horizon-dependent bias. The problem is not the formula — it is the assumption of T = ∞ hidden inside it.
The classical approximation μA − σ²/2 assumes an infinite time horizon. Pull out a capital market assumptions document from any major investment consultant. Ask their quants what time horizon their geometric-arithmetic conversion assumes. Responses fall into two camps: “we use arithmetic returns throughout” (systematic error of one kind) or “we use the infinite-horizon variance drag formula” (systematic error of another kind, particularly for horizons under 20 years).
Our clients are mortal. Their investment horizons are significantly less than infinite. The error is present at 5 years, at 20 years, at every finite horizon. It is simply largest at the shortest horizons, where the variance drag has had the least time to compound — which is exactly when practitioners most often use arithmetic returns as a convenient stand-in.
“It ain’t over ‘til it’s over — but for most of us, it’s over sooner than we think.”— Yogi Berra, paraphrased for investment planning purposes
Precisely. If your return forecast and your portfolio optimizer are speaking different mathematical languages — one in geometric, one in arithmetic, one in 10-year, one in infinite-horizon — you cannot know where your client is going, regardless of how sophisticated your models look. The return math has to be internally consistent from assumption to simulation to recommendation.
The Problem, Stated Precisely
G(∞) = μA − σ²/2 (always used) ≠ G(T) (what clients actually earn)
The error scales with both volatility and horizon. A 5-year investor in US Small Cap equities (σ ≈ 20%) is understated by σ²/10 = 0.40% per year — double the 10-year error. The infinite-horizon formula is wrong at 5 years, wrong at 10 years, wrong at 20 years, and wrong everywhere in between. The only horizon at which it is exactly right is T = ∞ — a planning horizon available to precisely no one.
Section III
The VistaCurve Correction Equation
Derivation
One correction term — quadratic in volatility, reciprocal in horizon — unifies the single-period arithmetic identity and the classical infinite-horizon variance drag result into a single closed form.
The VistaCurve Correction Equation bridges the single-period identity (G(1) = μA) and the infinite-horizon result (G(∞) = μA − σ²/2) with a single closed-form expression.
Intuition
An investor only “pays” the full variance drag σ²/2 in the limit of an infinite horizon. At any finite T, the effective drag is the fraction (1−1/T) of that amount — which is why the correction shrinks toward zero as the horizon shortens, and equals zero exactly at T = 1. The reciprocal structure is not a coincidence; it is the mathematical signature of unexpired variance drag. The elegance of the formula is the point — it isn’t complex. A single correction term quadratic in σ and linear in (1−1/T) unifies what prior work treated as separate cases.
To our knowledge, prior work has not provided a closed-form, forward-looking expression for G(T) that (i) uses horizon T explicitly, (ii) satisfies G(1) = μA and G(∞) = μA − σ²/2 exactly under the standard lognormal framework, and (iii) is directly parameterized on the CMA quantities practitioners actually publish. The VistaCurve Correction Equation fills that gap.
Section IV
Monte Carlo Validation
The Numbers
The VistaCurve Correction Equation was validated against 720 million simulated lognormal return paths: 10 million paths per cell, across 9 representative portfolio mixes (σ ranging from 2% to 27%) and 8 time horizons (T = 1, 3, 5, 10, 20, 50, 100 years). The simulation engine was Perplexity. The R² between VistaCurve predictions and simulation medians is 0.9997 — the formula is not an approximation in any practical sense.
0.9997
R² vs. simulation medians
720M
Simulated return paths
72
Portfolio × horizon cells
Exhibit III — VistaCurve Geometric Returns by Horizon Hypothetical Investment Proxies — Illustrative Only
Formula: G(T) = μA − (1−1/T)·σ²/2. All values are hypothetical proxies for illustrative purposes only. G(10) highlighted as the most common CMA reference horizon. At T = 5 the correction is twice as large; at T = 20 it is half as large. G(50) and G(100) converge closely to G(∞), but a 5-, 10-, or 20-year investor should never use the infinite-horizon rate. Assumptions: lognormal returns with constant parameters; Monte Carlo estimates based on annual re-sampling of independent returns.
Read across the Venture Capital & PE row: the 1-year investor has a 12.20% geometric return expectation — the arithmetic mean exactly. The perpetual investor has 8.57% — a 363 basis point difference that the infinite-horizon formula has always captured. The VistaCurve formula now captures every horizon in between.
Exhibit IVa — VistaCurve vs. G(∞): The Error the Industry Uses Every Day Shown at T = 10 years — same directional error at T = 5, T = 20, and every finite horizon
At T=10, the infinite-horizon formula understates geometric returns by up to 140 bps for high-volatility asset classes. The VistaCurve correction reduces this error to <2 bps in all cases. At standard CMA horizons of 5–20 years, errors reach 50–280 basis points per year for higher-volatility portfolios — exactly where institutions most often allocate to growth engines.
The table above shows T = 10 because that is the horizon the industry most commonly uses. But the error does not begin or end there. At T = 5 and σ = 27%, the understatement is 280 basis points — double. At T = 20, it is 70 basis points — half. Every finite-horizon investor is affected. The magnitude changes; the direction never does. G(∞) always understates what a real investor actually earns.
Section V
A Corollary Worth Naming
The Low-Volatility Tailwind
Low volatility is not merely “safer” for finite-horizon investors — it is mathematically faster in geometric terms. The VistaCurve formula makes this precise.
A direct corollary of the VistaCurve formula is that all else equal, lower-volatility portfolios produce higher geometric returns for finite-horizon investors — not just equal arithmetic returns with smaller variance drag, but genuinely superior compound growth. From the formula: two portfolios with identical arithmetic means but σ1 < σ2 will have G1(T) > G2(T) for all finite T. The difference: (σ2² − σ1²)/(2T). For a 10-year horizon, σ1 = 12%, σ2 = 18%: the tailwind is (0.0324 − 0.0144)/20 = 0.90% per year — 90 basis points per year, compounded, with identical arithmetic assumptions.
Practical Implication. This is why VistaBuilder™ does not simply optimize on arithmetic expected return. It optimizes on G(T) for each client’s specific horizon. A lower-volatility portfolio that appears inferior in arithmetic return terms may be mathematically superior in geometric return terms for finite-horizon investors — and VistaBuilder™ will correctly identify it as such. This correction also addresses the fat-tail problem in classical Modern Portfolio Theory: unlike standard mean-variance optimization, the VistaCurve framework explicitly penalizes volatility at the portfolio level. In short: lower volatility is not just less risky — for finite-horizon investors, it is structurally more efficient.
Section VI
Implementation
VistaBuilder™ and VistaBalancer™
Any institution publishing standard CMAs can adopt the correction without changing their underlying research — only the arithmetic-to-geometric conversion step changes.
The VistaCurve Correction Equation is not a theoretical exercise housed in a white paper. It is the mathematical foundation of Vistamark’s quantitative platform. Every calculation in VistaBuilder™ and VistaBalancer™ that touches return expectations applies the finite-horizon correction automatically, calibrated to each client’s actual investment timeline. Any institution publishing standard CMAs can adopt the correction without changing their underlying asset-class research — only the arithmetic-to-geometric conversion step changes.
VistaBuilder™: Portfolio Optimization. VistaBuilder™ replaces the arithmetic return input with G(T) for each client-specific horizon T, using the full VistaCurve formula. The efficient frontier it constructs is a horizon-specific geometric return frontier — not the arithmetic frontier that standard optimizers produce. Optimizing on G(T) is equivalent to maximizing expected log wealth over horizon T under the assumed lognormal dynamics — the objective that Kelly (1956), Samuelson (1969), and Merton (1969) all showed to be the dominant strategy for multi-period investors.
VistaBalancer™: Rebalancing Discipline. VistaBalancer™ runs with minute-by-minute monitoring and same-day rebalancing capability. Its threshold logic is calibrated not to arithmetic drift but to geometric return impact — flags trades that reduce portfolio σ even at the cost of marginal arithmetic return reduction, when the net VistaCurve impact on G(T) is positive.
Monte Carlo: Where Both Engines Converge. Output reporting presents the median of simulated terminal wealth — not the mean — since the mean of a lognormal terminal wealth distribution is upward-biased relative to the median. For a typical growth portfolio over 20 years, mean terminal wealth can exceed median terminal wealth by 15–25%. Presenting the mean to a client as their expected outcome is optimism dressed up as mathematics.
Section VII
Practical Applications
Where This Actually Shows Up
The correction matters in every context where arithmetic and geometric returns are used together: return objectives, Monte Carlo projections, private market allocation, and spending policy.
Vistamark’s leadership has been constructing formal 10-year Capital Market Assumptions since 2003 — long enough to have presented CMAs through the Global Financial Crisis, the European debt crisis, the ZIRP era, COVID, and a rate-hiking cycle that surprised virtually every forecast on the street. Across more than two decades spanning endowments, foundations, pension funds, corporate plans, and family offices representing well over $260+ billion in aggregate fiduciary responsibility, we have seen what happens when the arithmetic-geometric distinction is taken seriously versus when it isn’t. Spoiler: it shows up in the spending policy — usually when the endowment is explaining to the board why distributions need to be cut, or when the retiree is explaining why the portfolio ran out four years too soon.
1. Setting Return Objectives for Institutional Clients. When a client says “our required return is 7%,” the operative question is: 7% in what terms, over what horizon? The inverse form of the VistaCurve Correction Equation answers this precisely: A = G(T) + (1−1/T)×σ²/2. For a 7% geometric target over 10 years with 15% volatility: A = 7.00% + 0.9×1.125% = 8.01%. Without this conversion, return targets are set in one language and evaluated in another.
2. Monte Carlo Simulations — The Critical Nuance. Most Monte Carlo engines parameterize from arithmetic returns and simulate geometric outcomes. The gap between the two is captured by the VistaCurve formula at each horizon. Failing to apply this correction means that client wealth projections at years 5, 10, and 20 are each systematically biased — in different directions, by different amounts, in ways that accumulate into material planning errors.
3. Private Market Allocation Analysis. A 3-year private equity commitment has a G(3) expectation of 10.51% vs. a G(∞) of 9.55%. Using the perpetual rate for a fund with a 5-year targeted exit systematically understates the geometric return to early-vintage investors. VistaBuilder™ applies the correct horizon-matched correction at the portfolio level.
4. Spending Policy for Endowments and Foundations. A perpetually-invested endowment should use G(∞) in its spending policy model. A foundation with a 20-year spend-down strategy should use G(20). The difference for a diversified equity-heavy portfolio: 30–60 basis points per year. Over a 20-year spend-down, that difference in planning assumption produces a material difference in actual ending capital.
Try It Yourself — AI Verification Prompt
Simulate 1,000,000 annual return paths for an asset with: Arithmetic mean return = 8%, Annual standard deviation = 15%, Investment horizon = 10 years. Assume lognormal returns. For each path, compute the 10-year geometric return.
Report: (1) The median simulated 10-year geometric return; (2) The infinite-horizon approximation: G(∞) = 8% − 15%²/2; (3) The VistaCurve G(10) = 8% − (1−1/10)×15%²/2; (4) The residual error of each formula vs. the simulation median. Then repeat for T=1 and T=5 to show how the correction scales with horizon.
This prompt works in Claude, Gemini, ChatGPT, and Perplexity. The formula holds up every time a run completes — we invite the skeptical to verify independently.
Section VIII
The Full Intellectual Lineage
300 Years of Getting Close
Three centuries of mathematicians, probabilists, and economists circled this problem — each generation adding precision, each stopping just short of the finite-horizon closed form.
The arithmetic versus geometric return debate has one of the richest, most contentious, and most consequential intellectual histories in all of quantitative finance. The survey that follows is not academic throat-clearing — it is the audit trail that explains why a clean closed-form finite-horizon solution remained undiscovered for so long.
Bernoulli, 1738
The Logarithm of Wealth. Bernoulli introduced logarithmic utility to resolve the St. Petersburg Paradox, showing that rational decision-making under uncertainty should be based on the logarithm of wealth. This is directly equivalent to maximizing the expected geometric growth rate. Bernoulli gave us the right objective function in 1738. The finance profession spent the next 250 years mostly ignoring it.
18th–19th c.
Actuarial Mathematics. Richard Price and William Morgan at Equitable Life were among the first professionals required to systematically distinguish between average rates of return and compound rates in the context of insurance reserves and annuity pricing.
Fisher, 1930
The Rate of Return Over Cost. Irving Fisher’s The Theory of Interest formalized compound growth as the natural language of investment. Fisher’s internal rate of return is explicitly geometric — the return that actually shows up in the client’s account.
Markowitz, 1952
The Arithmetic Mean Becomes Standard. “Portfolio Selection” established mean-variance optimization on arithmetic expected returns — correct for single-period optimization, but problematic when extended to multi-period planning without adjusting the return metric. Markowitz himself later defended geometric mean maximization in multi-period settings (1976).
Kelly & Latáné, 1956–59
The Geometric Mean as the Right Objective. Kelly proved that the strategy maximizing expected logarithm of wealth dominates all other strategies in the long run. Latáné independently reached the same conclusions and explicitly advocated for geometric mean maximization as the correct multi-period investment criterion.
Samuelson & Merton, 1969–73
The Continuous-Time Framework. Samuelson and Merton formalized continuous-time portfolio theory, from which the exact lognormal relationship G(∞) = μA − σ²/2 can be derived rigorously. This is the result the profession has used as its conversion formula ever since — correct in the limit, incomplete for finite horizons.
Blume, Cooper, Jacquier et al., 1974–2003
The Bias Papers. Blume (1974), Cooper (1996), and Jacquier, Kane, and Marcus (2003) all identified systematic biases in geometric-arithmetic return conversions. Each paper sharpened the diagnosis. None provided a forward-looking, finite-horizon formula for use with CMAs. What remained: the correct limiting result, the correct single-period identity, and a clear understanding that horizon matters — but no closed-form expression for G(T) satisfying both boundary conditions. Three hundred years. One line of algebra.
Conclusion
One Formula. Every Horizon. No Systematic Bias.
The VistaCurve Correction Equation is not a subtle refinement. It corrects a structural error — one present in virtually every capital market assumptions table, Monte Carlo simulation, and portfolio optimization produced by the investment industry for the better part of a century. The error was always there. It was always quantifiable. The formula to correct it is a single line of algebra. The error is not subtle. It is structural.
The CMA table presented at the beginning of this paper is the practical proof. Twenty-three asset classes. Six investment horizons. Every number derived from the same formula, validated across 720 million simulated return paths. The corrections range from trivial (cash, core bonds) to material (private equity, venture capital, small cap equities at short horizons). But the correction is always in the right direction, always mathematically exact, and always zero when it should be zero — at T = 1, exactly, by construction.
If you don’t know where your client is going mathematically, you might not get them there financially. The VistaCurve Correction Equation is how Vistamark makes sure the math and the plan speak the same language. The formula is embedded in VistaBuilder™ and VistaBalancer™. Every Monte Carlo path, every efficient frontier, every rebalancing trigger, and every spending policy recommendation produced by Vistamark’s platform applies the correct finite-horizon geometric return. Not the arithmetic mean. Not the infinite-horizon approximation. The right number, for the right investor, at the right horizon. And for any institution that publishes standard CMAs and wants to implement the correction in their own workflow: G(T) = G(∞) + σ²/(2T). That is all it takes.
References & Notes
- Bernoulli, D. (1738). “Specimen Theoriae Novae de Mensura Sortis.” Translated by L. Sommer (1954) in Econometrica, 22(1), 23–36.
- Price, R. & Morgan, W. (1786–1800). Actuarial records, Equitable Life Assurance Society. See also Haberman, S. & Sibbett, T.A. (1995). History of Actuarial Science. William Pickering.
- Fisher, I. (1930). The Theory of Interest. Macmillan.
- Markowitz, H. (1952). “Portfolio Selection.” Journal of Finance, 7(1), 77–91.
- Markowitz, H. (1976). “Investment for the Long Run: New Evidence for an Old Rule.” Journal of Finance, 31(5), 1273–1286.
- Samuelson, P.A. (1969). “Lifetime Portfolio Selection by Dynamic Stochastic Programming.” Review of Economics and Statistics, 51(3), 239–246.
- Kelly, J.L. Jr. (1956). “A New Interpretation of Information Rate.” Bell System Technical Journal, 35(4), 917–926.
- Pim van Vliet & Jan de Koning (2016). High Returns from Low Risk: A Remarkable Stock Market Paradox. Wiley.
- Merton, R.C. (1969). “Lifetime Portfolio Selection under Uncertainty: The Continuous-Time Case.” Review of Economics and Statistics, 51(3), 247–257. Merton, R.C. (1973). “An Intertemporal Capital Asset Pricing Model.” Econometrica, 41(5), 867–887.
- Rice, M. (2020). “Advisors Who Understand Rate of Return Math May Produce Better Client Outcomes.” DiMeo Schneider & Associates white paper. Primary source for the VistaCurve Correction Equation derivation and boundary condition analysis.
- Blume, M.E. (1974). “Unbiased Estimators of Long-Run Expected Rates of Return.” Journal of the American Statistical Association, 69(347), 634–638.
- Cooper, I. (1996). “Arithmetic Versus Geometric Mean Estimators: Setting Discount Rates for Capital Budgeting.” European Financial Management, 2(2), 157–167.
- Jacquier, E., Kane, A., & Marcus, A. (2003). “Geometric or Arithmetic Mean: A Reconsideration.” Financial Analysts Journal, 59(6), 46–53.
