Projection vs Prediction
The most consequential error in long-term financial planning isn't mathematical—it's epistemological. It's confusing a prediction (a specific claim about what will happen) with a projection (a scenario showing what could happen under stated assumptions). This confusion leads investors and retirees to plan as though the future is knowable when it isn't, and then to panic when inevitable deviations occur.
When your financial advisor says "You'll earn 7% annually, accumulate $2.1 million by retirement, and withdraw $84,000 yearly," they're making predictions disguised as projections. They're stating a single point outcome as though it's certain. But markets don't deliver single points. They deliver distributions of outcomes, with wide uncertainty bands.
The difference between projections and predictions determines whether you build a resilient 30-year plan or a fragile one that breaks when the world deviates from a single assumed path.
Prediction: The Illusion of Certainty
A prediction is a claim about a specific future outcome. "The S&P 500 will return 9% next year." "You'll retire at 62 with $2 million." "Gold will reach $2,500 by 2027." These are predictions. They specify a single point in time and a single outcome.
The problem is that predictions about complex systems with many variables are notoriously unreliable. Financial markets aren't mechanical systems with predictable inputs. They're emergent systems where psychology, policy, technology, and countless unknowns interact. The future isn't a single timeline. It's a distribution of possible timelines.
Research on prediction accuracy in finance is sobering. A meta-analysis of stock market forecasts from 1973-2015 found that expert predictions (from professional economists and strategists) had virtually zero predictive power over actual returns. Whether experts predicted 5% or 12% returns, the actual returns were uncorrelated with their forecasts. They were no better than random guessing.
Yet financial advisory is built almost entirely on predictions. Advisors typically provide a single projected return (often 7-8% for stocks), a single assumed life expectancy, a single assumed inflation rate, and a single projected portfolio outcome. The client sees one number and assumes that's the plan.
This creates a psychological setup for failure: when actual outcomes deviate (and they always do), the plan feels broken. A client who was shown "$2.1 million by retirement" and accumulates $1.8 million or $2.4 million might feel disappointed in the $1.8 million scenario, even though $1.8 million is a successful outcome. The prediction set false expectations about certainty.
Projections: Planning with Uncertainty
A projection, by contrast, acknowledges uncertainty explicitly. Instead of "You'll earn 7% and accumulate $2.1 million," a projection says: "Under a 7% annual return assumption, you'll accumulate $2.1 million. Under 5% annual returns, you'll accumulate $1.6 million. Under 9% annual returns, you'll accumulate $2.9 million. Historical experience suggests actual returns will fall somewhere in this range, most likely between 5% and 9%."
This is a projection—a scenario showing outcomes across different assumptions. It acknowledges that the future is uncertain and provides a range rather than a point. A good projection includes multiple scenarios (low, base, high) and explicitly states which assumptions might prove wrong.
A sophisticated projection goes further, using probability distributions rather than discrete scenarios. Research from the Federal Reserve on economic projections and analysis from academic finance shows this approach: instead of three scenarios, hundreds of simulations across uncertainty ranges produce a probability distribution of outcomes.
For a retiree, this might show: "There's a 70% probability your portfolio will support your planned spending for 40 years. There's a 15% probability it will be exhausted by year 35. There's a 15% probability you'll have surplus funds." This honest framing acknowledges reality: you can't be certain the plan will work, but you can quantify the probability.
The difference is profound. Predictions create false confidence. Projections create informed realism. And realistic long-term plans actually work better than overconfident ones, because realism prompts you to build contingencies.
Why Predictions Feel Better Than Projections
Human psychology heavily favors predictions over projections. A prediction gives closure—you know the plan, you follow it, you arrive at a specific destination. A projection creates ambiguity—you know the range of outcomes but not which you'll experience. We're psychologically built to reduce uncertainty, so predictions feel more comfortable.
Additionally, projections require more cognitive effort to understand. "You'll have $2.1 million" is simple. "You'll likely have between $1.6 million and $2.9 million, with a 70% probability in the $1.8-2.4 million range" requires more mental processing. People naturally prefer simplicity.
Advisors often provide predictions instead of projections for these same reasons—clients prefer them, they're easier to present, and they sound more authoritative. But this comfort comes at a cost: fragility. When reality deviates from the single predicted path, the plan feels broken.
Consider a retiree who was told "Your portfolio will last 40 years." When markets decline 20% in year three and their sequence of returns calculation shows the portfolio will be exhausted in year 35 under historical return assumptions, they panic. They feel their plan has failed. But their plan always had a distribution of outcomes. In this scenario, they simply ended up in the lower tail of that distribution.
A retiree who was shown upfront that there's a 20% probability the portfolio won't last 40 years—and who built contingency plans around that reality (like flexible spending, part-time work options, or delayed retirement)—experiences the same market decline without the same panic. They were projecting, not predicting. They expected this possibility and planned for it.
The Math: Why Distributions Matter More Than Points
Here's the statistical reason projections matter more than predictions:
Suppose you assume 7% annual stock returns. If returns were certain at 7%, a $100,000 portfolio would grow to $761,000 in 30 years—a single point, fully predictable.
But actual returns are volatile. Historically, annual stock returns range from -37% to +54%. The average is 10% nominal (7% real after inflation), but actual years rarely deliver the average. Most years deliver 0-15%, some deliver negative returns, a few deliver extreme returns.
When you compound across 30 years with variable returns, you don't get the mathematically "expected" outcome. You get one realization from a distribution of possible outcomes. And here's the crucial part: due to the convexity of compounding, the most likely single outcome is actually worse than the average outcome.
This is called "volatility drag." A portfolio averaging 7% annual return with zero volatility compounds differently from a portfolio averaging 7% with 15% annual volatility. The volatile portfolio compounds slightly slower because losses in bad years hit a smaller base and gains in good years hit that smaller base.
More importantly, you don't experience the average. You experience one path through return space. A $100,000 portfolio might compound to $600,000 in one 30-year period and $900,000 in another, both with the same 7% average return—just different sequences. Over a single 30-year lifetime, you get one outcome, not the average.
This is why projection ranges matter. You can't know which outcome you'll experience. You can only quantify the probability distribution and plan for resilience across that distribution.
Research on retirement planning Monte Carlo simulations shows that plans accounting for return distribution and sequence-of-returns risk show success rates 10-20 percentage points higher than plans based on single-point assumptions, because they account for the actual complexity of real-world outcomes.
Practical Distinction: What This Means for Your Plan
In practical terms, here's how to distinguish predictions from projections in your planning:
Red flags for predictions:
- A single specific outcome ("You'll have $2.1 million")
- No acknowledgment of what could go wrong
- No discussion of uncertainty or variance
- Language suggesting certainty ("You will accumulate...")
Green flags for projections:
- A range of outcomes ("Most likely $1.8-2.4 million, could be as high as $2.9 or as low as $1.2")
- Discussion of scenarios and sensitivity (what if returns are lower? What if you live longer?)
- Explicit probability statements ("70% probability of success")
- Language acknowledging contingency ("You likely will, assuming...")
- Multiple return scenarios tested (5%, 7%, 9%)
A strong retirement plan includes projections like: "Assuming 6% real returns and a 3% spending rate, there's an 85% probability this portfolio lasts 40 years. Assuming 4% real returns, there's a 65% probability. If returns fall below 3%, spending flexibility or extended working years would be needed."
Notice what this does: it quantifies uncertainty, it identifies the threshold where the plan becomes fragile (3% returns), and it prompts contingency planning. You're not pretending the future is knowable. You're planning intelligently within known uncertainties.
The Mistake: Planning for a Single Future
Most financial plans implicitly assume a single future and fail when that future doesn't materialize. You plan for a 7% return, markets deliver 4.5%, and your plan is "broken." You plan to retire at 62, a health crisis forces retirement at 60, and your plan is "broken." You plan for a 30-year retirement, live to 95, and your plan is "broken."
But the plan wasn't broken. It was never designed for this outcome. You projected for one possibility and got another. This is inevitable—the future is uncertain, and any single projection will miss.
The solution is building projections robust across uncertainty ranges. Instead of projecting a single retirement date, project multiple dates (retire at 60, 62, 65) and identify the probability of comfortable retirement at each. Instead of one return assumption, test against return distributions. Instead of a single spending level, test different spending flexibility approaches.
A client who plans to retire on $100,000 annually, but whose projections show 85% success with that spending and 99% success with $80,000 flexible spending (dropping to $75,000 if markets are down), has a robust plan. They can retire with confidence, understanding the trade-off between spending flexibility and success probability.
Compare this to a client told "You'll retire at 62 with $100,000 annually." If markets are down when they retire, or if they live longer, the plan feels failed, even though a more robust planning approach would have revealed the risk upfront.
Using Projections for Decision-Making
Strong projections inform better decisions:
Comparing savings rates: Projection A (saving 15% of income) shows 95% success. Projection B (saving 20%) shows 99% success. The 4% improvement might not justify the 5 percentage point increase in savings rate. You can make an informed trade-off.
Evaluating early retirement: Projection shows that retiring at 55 instead of 60 drops success rate from 90% to 70%. You might be willing to work 5 more years for a 20 percentage point improvement. Or you might prefer retirement at 55 and implement flexible spending to manage the higher failure probability. Either way, you're making informed decisions.
Deciding between volatility levels: A conservative portfolio (40% stocks, 60% bonds) shows 85% success. An aggressive portfolio (80% stocks, 20% bonds) shows 90% success. You can decide whether the 5 percentage point improvement justifies 3x higher portfolio volatility.
Without projections, you're flying blind—guessing at which decisions matter. With projections, you're informed about trade-offs and can optimize deliberately.
The Role of Assumptions in Projections
An important caveat: projections are only as good as their assumptions. A projection assuming 3% real returns for 50 years in a 1.5% real-rate environment might be unrealistic. A projection assuming you never adjust spending when markets decline might misrepresent your likely behavior.
The key is examining assumptions explicitly. Good projections include sensitivity analysis: "If real returns are 2% instead of 3%, the success probability drops from 85% to 72%. If you adjust spending 15% downward during down markets, the success probability improves to 92%."
When reviewing projections—whether from an advisor or your own calculations—always ask: Which assumptions am I uncertain about? How sensitive are the results to those assumptions? What would change the conclusions?
For example, research from the Social Security Administration on life expectancy shows individual variation around average life expectancy. Planning for a single life expectancy (say, 90) is less robust than projecting across a range (85-95+) and understanding the implications.
Prediction vs. Projection Frameworks
Real-World Example: The Difference in Action
Two retirees, both retiring at 62 with $1.2 million portfolios and $60,000 annual spending targets:
Retiree A uses prediction: Advisor shows "$1.2M at 7% equals 30 years of $60k spending. You're good." Single point forecast. No discussion of variance.
Retiree B uses projections: Advisor shows "Assuming 6% real returns, there's 85% probability of 40-year success. Assuming 4% returns, probability drops to 65%. Assuming 7% returns, probability is 95%. Markets historically return 5-7% but can vary significantly. Your plan is most sensitive to spending flexibility—if you maintain $60k regardless of market returns, success drops to 70%. If you adjust spending 20% downward in down markets, success improves to 90%."
Retiree A feels certain. Retiree B understands the uncertainty.
In 2024, markets deliver 8% returns. Both retirees are pleased. In 2025, markets decline 15%. Retiree A sees the decline, checks their plan, and realizes market returns are now below their 7% assumption. They panic, thinking the plan is broken. They consider reducing spending immediately or working longer.
Retiree B sees the same 15% decline and checks their projections. The decline is within the expected range. They adjust spending 10% downward (well within the flexibility they planned for) and continue. They understand this single-year deviation doesn't break their long-term plan.
Retiree A's prediction created false confidence followed by panic. Retiree B's projections created informed confidence followed by systematic adaptation.
Summary
Predictions are specific claims about single future outcomes. Projections are scenario analyses showing ranges of outcomes under different assumptions. Financial plans based on predictions are fragile—they break when the future deviates from a single assumed path. Plans based on projections are robust—they acknowledge uncertainty and build contingencies accordingly.
For 30-year financial planning, projections are essential. Test your plan across multiple return scenarios, spending flexibility options, and longevity timeframes. Understand probability distributions rather than point forecasts. Build contingencies for the scenarios that matter most.
The future isn't knowable, but it is quantifiable in probability terms. Projects embrace this reality. Predictions ignore it, and pay the cost in fragility. A robust 30-year plan projects across uncertainty. It acknowledges what could go wrong, quantifies the probabilities, and builds contingencies. That's how you plan intelligently for a future you cannot predict.