Investment Portfolio Generator
Describe your goal, set your risk profile, and CleaRank Financial AI designs the portfolio for you. The generator runs six frontier optimizations in parallel, pulls real historical price data for every recommended holding, runs a 1,500-path Monte Carlo against a Cholesky-decomposed correlation matrix, stress-tests the basket against 2008, 2020 and 2022, and delivers an allocation with the eight statistics that matter (Total Return, Annualized, Volatility, Sharpe, Sortino, Max Drawdown, Beta, Alpha). Free, no signup, no upload. The sibling of the Portfolio Visualizer: visualizer backtests baskets you already own, the generator designs baskets you do not yet have.
How the portfolio generator works
Six steps, one click. Fill in the amount, period, risk profile and strategy. Drop a sentence or two into the goals box about what you are actually trying to do with this money. Click Generate. CleaRank Financial AI picks the candidate universe, the six-stage backtest pipeline pulls real historical data, six frontier optimizations run in parallel (Min Variance, Max Sharpe, Max Return, Max Information Ratio, Risk Parity, Hierarchical Risk Parity), and the generator returns a recommended portfolio along with five alternative variants you can switch between with one click. Every chart and statistic below is computed from real daily closes, not modelled from theoretical inputs.
1. Tell CleaRank your amount, period, risk and strategy
The top of the form is a single row of four controls. Investment Amount sets the starting capital with a default of $10,000, which the recommended allocation grows or shrinks bar by bar through the analysis period. Period sets the analysis horizon (default 5 years) and dictates how much historical data the backtest pulls for every recommended holding. Risk Profile is the five-step dial that controls everything downstream: Low, Low-Moderate, Moderate, Moderate-Aggressive, Aggressive. Each step maps to a target volatility band and a target maximum drawdown ceiling, which the optimizer treats as hard constraints. Strategy is the universe selector with six options: Balanced Growth, Aggressive Growth, Conservative Income, Dividend Growth, Tech Innovation, Global Diversified. The strategy seeds the candidate set of holdings (US large-cap, international, bonds, dividend aristocrats, tech megacaps, etc.) and the risk profile decides how to weight them.
2. Describe what you want in plain English
The goals textarea is where the generator stops being a checkbox builder and starts behaving like a financial advisor. Type up to 500 characters describing what this money is actually for. A retirement nest egg with a 20-year horizon. A house deposit in 3 years. A dividend-income sleeve targeting 4 to 6 percent yield. A taxable account with a high tax bracket where capital gains efficiency matters. The generator reads the text and adjusts the candidate universe: it tilts toward tax-efficient ETFs when you mention taxable, leans into long-duration assets when you mention 20-year horizon, biases toward dividend aristocrats when you mention income. Below the textarea, three optional fields refine the request further. Target amount turns on the live Goal Feasibility card: type $250,000 and the card tells you the required CAGR (9.2% in the screenshot) and grades it as Comfortable, Moderate, Aggressive, or Very Aggressive based on how close it is to historical mean-reversion bands. Monthly DCA contribution tells the generator how much fresh capital flows in every month, which materially changes the recommended allocation (dollar-cost averaging favours higher-volatility holdings). Max positions caps the basket size between 6 and 30 holdings (default 20), Leverage OK permits leveraged ETF candidates if checked, and Include factor tilts (on by default) lets the optimizer tilt toward value, quality, momentum or low-vol factor ETFs where the data justifies it.
3. Watch CleaRank Financial AI run six frontier portfolios in parallel
Click Generate and the same six-stage pipeline that powers the Portfolio Visualizer fires up. Stage one, asking CleaRank Financial AI for picks based on your strategy and goal text. Stage two, fetching real historical market data for every recommended ticker plus the benchmark. Stage three, computing daily returns and volatility. Stage four, running 1,500 Monte Carlo paths against the Cholesky-decomposed covariance. Stage five, building the correlation matrix. Stage six, synthesising the plain-language analysis from CleaRank Financial AI. What makes the generator different from the visualizer is the six-variant grid the pipeline produces. Instead of one optimal portfolio, the generator returns the entire efficient frontier as six concrete portfolios you can switch between in one click: Min Variance (the safest practical mix), Max Sharpe (best risk-adjusted return, the default), Max Return (the upside swing), Max Information Ratio (most alpha vs the benchmark), Risk Parity (equal risk contribution per asset), and Hierarchical Risk Parity (a tree-clustered version of risk parity that handles correlation regimes better). Pick the variant whose verdict matches your goal.
4. Read the allocation, the backtest, and the eight stats that matter
The first card the generator renders is the hero allocation: a conic pie chart of every recommended holding with weight, plus four meta pills (amount, period, rebalance frequency, expected yield) and a one-paragraph thesis written by CleaRank Financial AI. Below the pie sits the allocation table with ticker, name, sector, weight and dollar amount per holding. The free embed shows the top 5 holdings by weight (which is usually 70 to 85 percent of total capital). The remaining positions sit behind the Ultra Full Allocation Breakdown lock card. Underneath the table, the eight statistics that matter render in a grid: Total Return over the analysis period, Annualized Return, Annual Volatility, Sharpe Ratio, Sortino Ratio (penalises downside volatility only), Max Drawdown, Beta vs benchmark, and Alpha vs benchmark. Right after the stats grid the generator renders the historical backtest chart: the recommended portfolio as a pink equity curve against SPY in teal, day by day across the analysis period. Crisis windows show up on sight, so you can see exactly when the recommended portfolio was right, when it was wrong, and by how much.
5. Stress-test the generated portfolio against real market crises
Three diagnostics turn a one-line recommendation into a real risk model. The Monte Carlo 3-band projection renders 1,500 simulated forward paths against a Cholesky-decomposed correlation matrix, then plots the p10, p50 (median) and p90 final-balance bands across the analysis period. The width between p10 and p90 is exactly how much variance to expect from the recommended portfolio looking forward, even if the average return is good. The Correlation Matrix renders a heatmap of pairwise correlations between every holding plus a diversification ratio (the volatility you saved by combining versus the weighted sum of standalone vols). The matrix is free in the generator (it sits behind Ultra in the visualizer, free here because it is core to understanding the recommended basket). The Historical Stress Tests card replays the recommended portfolio through the 2008 financial crisis, the 2020 COVID drawdown, and the 2022 stocks-and-bonds correlated drop, showing what the recommended basket would have done in each window versus SPY. Alongside the stress tests sit the Advanced Risk Metrics card (CVaR 95 one-day, downside deviation, beta-adjusted alpha) and the Market Regime card (current VIX, Fear & Greed, Shiller CAPE, CAPE-adjusted 10-year forward return).
6. Unlock the full AI rationale and every holding
Two diagnostics ship locked on the free embed and unlock together with an Ultra subscription. AI Portfolio Rationale writes a per-holding thesis: “VTI: core US equity sleeve, the growth engine, selected over SPY for broader small and mid-cap exposure”. “VXUS: geographic diversification, decorrelates during US-led drawdowns”. “BND: ballast for the drawdown, the 2022 correlation breakdown is acknowledged but bonds remain the strongest mainstream diversifier across longer horizons”. The rationale also explains how the engine balanced risk versus return, why this risk profile favoured these tilts, and writes the rebalancing guidance you should follow if the basket drifts from target weights. The free embed shows a 280-character teaser of the rationale. Full Allocation Breakdown is the second Ultra-locked card. The free embed shows the top 5 holdings (70 to 85 percent of capital). Ultra unlocks all N positions in the recommended basket, lets you place all N orders straight into the paper-trading Simulator with one click, lets you compare four portfolios side by side, and adds custom-blend scenarios you can save and reload. The generator renders blurred mockups in place of the live cards so you can see exactly what you are missing before you upgrade.
Built for investors who want a designed portfolio, not a guessed one
First-time ETF investors, retirement planners modelling a 30-year horizon, active traders looking for a passive sleeve, advisors generating client proposals. Same six-variant frontier output, four very different starting points. Pick the workflow that matches yours.
First-time ETF investors building their first taxable portfolio
You have $10,000 in a brokerage account and no idea where to start. Pick Moderate risk, Balanced Growth strategy, type “first taxable account, 20-year horizon, broad-market exposure” into the goals box. The generator returns six concrete allocations with real ETFs (VTI, VXUS, BND, VNQ, GLD) and the eight statistics for each. Switch to Min Variance for the safest mix, or Max Sharpe (the default recommendation) for the best risk-adjusted return.
Retirement planners modelling a 30-year horizon
Set the target amount field to your retirement number. Period 10y. Monthly DCA to your current contribution. Type “retirement nest egg, 30-year horizon, comfortable with 25 percent drawdowns” into the goals box. The Goal Feasibility card grades whether your target is realistic. The Monte Carlo p10 path tells you whether the bad-case scenario still gets you there. If p10 misses, raise the monthly DCA until it lands.
Active traders looking for a passive sleeve
You run an active book but want a passive sleeve for the capital you do not want to risk on the next trade. Set Aggressive Growth strategy, Moderate-Aggressive risk, period 10y, “passive sleeve to balance my active trading book” in goals. The generator returns a 6-asset basket with real correlations against your active book, plus the Risk Parity variant for true equal-risk-contribution weighting. Send the result to the Visualizer to test it against your active P&L history.
Advisors generating client proposals quickly
Generate three proposals in five minutes, one for each risk profile (Low-Moderate / Moderate / Moderate-Aggressive). Same strategy, same period, same target amount. The variant grid lets a client pick between Min Variance for the cautious side, Max Sharpe for the default, Max Return for the aggressive side. Export the resulting allocation, drop the pie chart and stats grid into your IPS document, ship the proposal.
Why this portfolio generator runs deeper than the rest
Most free portfolio generators on the open web stop at a risk questionnaire and a pie chart of three ETFs. The CleaRank version returns the entire efficient frontier as six concrete portfolios, then backs each one with the same real-data pipeline that powers the visualizer. Six frontier portfolios, not one. Min Variance for the safest mix, Max Sharpe for the best risk-adjusted return (default), Max Return for upside, Max Information Ratio for benchmark-beating alpha, Risk Parity for equal-risk-contribution weighting, and Hierarchical Risk Parity for a tree-clustered version that handles correlation regimes better. Switch between them with one click without re-running the backtest. Real historical backtests, not Markowitz toy assumptions. Every metric in the eight-stat grid is computed from real daily closes pulled across your chosen period. 1,500-path Monte Carlo with Cholesky-decomposed correlation, so forward projections respect how your recommended holdings actually co-move. Three real historical stress tests: 2008, 2020 COVID, and 2022 inflation, replayed on the recommended basket. Factor tilts (value, quality, momentum, low-vol) selectable as a checkbox. Goal feasibility grading live as you type the target amount. No signup, no upload, no tracking.
The generator is the front door of a two-tool workflow. After the generator designs the portfolio, three cross-tool actions ship inside the result panel. Deploy to Simulator pushes all N recommended orders into the paper-trading Simulator with one click (Ultra unlock for the full all-N deploy; the free embed deploys the top 5). Send to Visualizer hands the recommended basket over to the Portfolio Visualizer for a deeper backtest with a different benchmark, a longer window, or custom weight tweaks. Save to Ideas drops the basket into the persistent Portfolio Ideas shelf alongside the 13 curated portfolios from the 543 indexed fund managers. The three actions form a single workflow: design with the generator, validate with the visualizer, paper-trade with the simulator, archive with ideas. Pro and Ultra subscribers get the same generator inside the full 22-tool workbench at trade.clearank.com, with the AI Portfolio Rationale and Full Allocation Breakdown unlocked.
What is a portfolio generator, in plain English
A portfolio generator is a tool that takes a goal and a risk profile as input and returns a recommended basket of investments as output. You describe what you want (capital preservation, retirement nest egg, dividend income, aggressive growth). The generator picks a candidate universe based on the strategy you select, then runs an efficient-frontier optimization that finds the weights that produce the best risk-adjusted outcome inside your stated risk band. The output is a list of holdings with target weights, plus the eight statistics (Total Return, Annualized, Volatility, Sharpe, Sortino, Max Drawdown, Beta, Alpha), plus a Monte Carlo projection of how the basket might evolve forward, plus stress tests against historical crises. In one screen you get the same kind of allocation a fee-based advisor would build in a Monday afternoon meeting, except backed by real backtests rather than asset-allocation lookup tables.
The reason serious investors use portfolio generators (rather than picking ETFs from a top-10 list) is that asset selection is the smallest part of the problem. The big part is position sizing: how much capital goes into each holding so that the basket sits at your target volatility, your target max drawdown, and your target Sharpe ratio. Getting position sizing right means solving a convex optimization across the historical covariance matrix of your candidate assets, which is a 30-minute computation that no human eyeballs. The generator runs the optimization six different ways, surfaces all six results, and lets you pick the variant whose risk-return trade-off matches the goal you typed into the form.
The generator is not the same thing as a robo-advisor. A robo-advisor picks a model portfolio off a lookup table based on your answers to a 10-question quiz, then sells you the management of that portfolio for an annual fee (typically 25 to 50 basis points). The CleaRank generator does the same allocation work in your browser for free, returns the exact ticker list and weights, lets you switch between six variants instead of locking you to one, and never holds your money. You take the recommended allocation, open a brokerage account anywhere, and place the orders yourself. Or you push the basket into the paper-trading Simulator first and watch how it behaves for a few weeks before committing real capital. Either way, you keep custody of your own money and you avoid paying recurring management fees on capital you allocated yourself.
“Allocation is the only investment decision that compounds. The single stock you pick matters less than the percentage you put into it. A portfolio generator solves that math once, surfaces the answer, and lets the rest of the investing process be boring. Boring is the goal. Boring is what compounds.”
The decision tree behind every generated portfolio
The generator runs a four-step decision tree the moment you click Generate. Each step takes a single input from the form and converts it into a hard constraint on the optimization that follows. The list below is the exact order the pipeline executes.
- Risk Profile becomes a target volatility band. Low maps to 4-7% annual vol with a 10% max drawdown ceiling. Low-Moderate maps to 6-10% / 15%. Moderate maps to 9-14% / 22%. Moderate-Aggressive maps to 13-18% / 30%. Aggressive maps to 17%+ / 40%. The optimizer treats those bands as hard constraints: portfolios outside the band get rejected before they ever reach the variant grid.
- Strategy seeds the candidate universe. Balanced Growth pulls broad-market ETFs (VTI / VXUS / BND / VNQ / GLD). Aggressive Growth tilts toward high-beta growth (QQQ / VUG / SCHG). Conservative Income tilts toward bond ETFs and dividend aristocrats (BND / VYM / SCHD / TLT). Dividend Growth narrows further to dividend-grower ETFs. Tech Innovation pulls megacap tech and disruptor ETFs. Global Diversified pulls geographic and asset-class diversifiers.
- CleaRank Financial AI reads the goals textarea and tilts the universe further. Mention “tax-efficient” and the universe biases toward ETFs over mutual funds. Mention “20-year horizon” and the universe accepts longer-duration and higher-vol candidates. Mention “income” and it narrows to dividend-payers and bond aggregates.
- Six frontier optimizations run in parallel. Min Variance minimises portfolio volatility subject to the universe and risk-band constraints. Max Sharpe maximises annualized return divided by volatility (the default recommendation). Max Return maximises annualized return subject to the max-drawdown ceiling. Max Information Ratio maximises alpha-per-tracking-error vs the benchmark. Risk Parity equalises the risk contribution of each holding. Hierarchical Risk Parity clusters the correlation matrix into a tree first, then risk-parity-weights across the tree.
Worked example: four goals, four generated portfolios
Same generator, four different inputs, four different recommended baskets. The top-3 holdings, Sharpe and Max Drawdown numbers below are illustrative ranges typical for each profile across a 5 to 10 year backtest against SPY as benchmark.
Four very different goals, four very different recommended portfolios. Notice the Conservative $50K basket posts the lowest Sharpe but the shallowest drawdown by a wide margin: that allocation is built to survive, not to maximise return. The Balanced $25K basket wins on Sharpe because the four-asset mix diversifies real risk rather than just adding tickers. The Aggressive $10K basket posts a Sharpe in the same band as Tech Innovation but with deeper drawdowns, because the risk is more diffuse rather than more efficient. The Tech Innovation basket has the highest beta and the most single-factor risk: it lives and dies on the megacap tech cycle. Same generator, four different verdicts. The generator does the math once and lets the goal you typed in determine which verdict applies to you.
Risk profile to typical allocation
Illustrative typical-allocation bands for each risk profile under the Balanced Growth strategy across a 10-year backtest against SPY. Use this as a sniff test for your own generated portfolio: if the result returns weights or stats far outside these bands, the strategy or the goal text is the reason. Equity % is the combined US plus international stock weight. Bonds % is the BND-style aggregate. Gold & Commodities % is the diversifier sleeve (GLD, DBC). Expected Return and Vol are annualized. Max DD is the typical peak-to-trough on a 10-year backtest.
The pattern jumps off the table. Each step up in risk profile trades a smaller bond ballast for a larger equity sleeve and reaches for higher expected return, with a directly proportional rise in volatility and max drawdown. The Sharpe ratio peaks somewhere in the Moderate to Moderate-Aggressive band because that is where the diversification benefit of multi-asset baskets is strongest. Aggressive baskets often post a lower Sharpe than Moderate ones even though their expected return is higher: that is the risk-adjusted price of reaching for upside.
| Profile | Eq % | Bnd % | Gld/Cmd % | Ann Rtn | Vol | Max DD |
|---|---|---|---|---|---|---|
| Low | 25% | 65% | 10% | +5.2% | 5.8% | -8% |
| Low-Mod | 40% | 50% | 10% | +6.8% | 7.6% | -13% |
| Moderate | 60% | 30% | 10% | +8.4% | 10.2% | -19% |
| Mod-Agg | 75% | 15% | 10% | +10.2% | 13.4% | -27% |
| Aggressive | 90% | 0% | 10% | +11.8% | 17.6% | -38% |
Illustrative bands. Run your exact goal text through the generator for a portfolio backtested against your real holding period and goal.
Five mistakes that make a generated portfolio fail you
A generator is only as honest as the inputs you give it. The five mistakes below turn a green Sharpe ratio into a red brokerage statement six months later. Each one has a one-line discipline that prevents it.
Picking Aggressive without the stomach for it
The Aggressive risk profile is built around a 40% max drawdown ceiling. That is a real number. When it happens, $10,000 becomes $6,000 on paper for months at a time. Investors who pick Aggressive on the form and then panic-sell at -25% printed the worst outcome possible: they wore the entire drawdown without getting any of the recovery. Pick the risk profile you can hold through the worst case, not the one that produces the prettiest projection.
Ignoring the goal feasibility card
The Goal Feasibility card grades whether your target amount is realistic at the implied CAGR. If it says “Very Aggressive: leverage or concentration”, that is the generator telling you the math does not work without taking on risk you said you did not want. Either raise the monthly DCA, extend the period, lower the target, or accept a higher risk profile. Do not generate a portfolio against a target the feasibility card already flagged red.
Setting Max Positions too low
The Max Positions field defaults to 20 for a reason. Setting it to 6 or 8 forces the optimizer to concentrate the basket into a handful of ETFs, which raises single-asset risk substantially. The Correlation Matrix and Risk Contribution diagnostics will show one or two holdings driving 50%+ of total portfolio variance. If you want a small number of holdings, accept the concentration risk and pick the Min Variance variant rather than Max Sharpe.
Generating once and never rebalancing
The generator returns target weights, not permanent weights. After 12 months the equity sleeve will have drifted (probably up, probably a lot), the bond sleeve will have drifted (probably down), and your real-world allocation will no longer match the recommended one. Rebalance quarterly or annually back to the target weights. Without rebalancing, a 60/40 portfolio drifts into an 80/20 portfolio over a 5-year bull market, and the drawdown you see in the next bear will be the 80/20 drawdown, not the 60/40 one you signed up for.
Confusing backtest results with future returns
The generator returns historical metrics. A 10.4% annualized return across the last 7 years does not mean 10.4% next year. Historical volatility, drawdown and correlation patterns are much stronger predictors of future risk shape (correlations often above 0.7) than historical returns are of future returns (correlations often below 0.4). Use the Monte Carlo p10 to p90 band as your real expectation, not the headline annualized number. The p10 path is the bad-case outcome, and that is the number to plan around.
Frequently asked questions
How does an AI portfolio generator differ from a robo-advisor?
A robo-advisor takes your money, picks a model portfolio off a lookup table based on a 10-question quiz, then charges 25 to 50 basis points annually to manage that money. The CleaRank portfolio generator does the same allocation work in your browser for free, returns the exact ticker list and target weights, and never holds your capital. Three concrete differences. First, the generator runs six frontier optimizations in parallel (Min Variance, Max Sharpe, Max Return, Max Information Ratio, Risk Parity, Hierarchical Risk Parity) instead of locking you to one model portfolio. Second, it backs every recommendation with a real historical backtest against daily price data plus 1,500-path Monte Carlo and three crisis stress tests, not just an asset-allocation lookup. Third, it never holds custody of your money: you take the allocation list, open a brokerage account anywhere, place the orders yourself, and rebalance on your own schedule. You keep the recurring management fee.
Which risk profile should I pick on the portfolio generator form?
Pick the risk profile you can hold through the maximum drawdown, not the one that produces the prettiest expected return. The five profiles map directly to drawdown ceilings the optimizer treats as hard constraints. Low: max drawdown 10%, target vol 4 to 7%, expected return roughly 5%. Low-Moderate: max drawdown 15%, target vol 6 to 10%, expected return roughly 7%. Moderate: max drawdown 22%, target vol 9 to 14%, expected return roughly 8 to 9%. Moderate-Aggressive: max drawdown 30%, target vol 13 to 18%, expected return roughly 10%. Aggressive: max drawdown 40%, target vol 17%+, expected return roughly 12%. If a 40% drawdown on $10,000 (real-world, $4,000 paper loss for many months) would force you to sell, do not pick Aggressive. The honest test is to picture the worst-case dollar number printed on a brokerage statement and ask if you would keep the position open.
What is the difference between the Aggressive Growth and Tech Innovation strategy?
The strategy selector decides which candidate universe the optimizer picks from. Aggressive Growth seeds a broad high-beta basket: QQQ, VUG, SCHG, large-cap growth, with a mix of small-cap growth and emerging-market exposure to diversify the bet across multiple growth factors. The basket is concentrated on growth as a style, but not on any single sector. Tech Innovation narrows the universe much further: QQQ, SMH (semiconductors), SOXX, XLK (technology select sector), ARKK (disruptive innovation), megacap tech individual names where the universe allows. The basket is concentrated on technology as a sector, with semis as a sub-tilt. Result: Tech Innovation is a single-sector bet with high beta and tight correlation to the megacap tech cycle. Aggressive Growth is a multi-sector growth-style bet with more diversification across factors. A Tech Innovation basket will outperform Aggressive Growth in a tech bull market and underperform sharply when tech corrects (2022 is the obvious example). Pick Tech Innovation only if you specifically want sector concentration.
Why does the generator return six frontier portfolios instead of just one?
Because “the best portfolio” depends entirely on what you are optimizing for, and the honest answer is that there are at least six legitimate optimization targets. Returning only one would lie about that. The six variants the generator runs in parallel are: Min Variance (minimises portfolio volatility, the safest practical mix), Max Sharpe (maximises annualized return divided by volatility, the default recommendation and the closest thing to a “best risk-adjusted return”), Max Return (maximises return subject only to the max-drawdown ceiling, for upside-seeking allocations), Max Information Ratio (maximises alpha per unit of tracking error vs the benchmark, for active-overlay strategies), Risk Parity (equalises the dollar risk contribution of each holding rather than the dollar capital weight), and Hierarchical Risk Parity (clusters the correlation matrix into a tree first, then risk-parity-weights across the tree, which handles correlation regimes better than standard risk parity). The variant grid lets you switch between all six in one click without re-running the backtest, so you can compare the trade-offs visually before committing.
Can I customise the generated portfolio or change the recommended weights?
Yes, in three ways. First, the variant grid lets you switch between Min Variance, Max Sharpe, Max Return, Max Information Ratio, Risk Parity and Hierarchical Risk Parity in one click without re-running the pipeline. Each variant produces a different weight mix from the same candidate universe. Second, the weight sliders under the allocation table let you manually adjust individual holding weights, with a live total-weight tracker that flags when you drift away from 100%. The eight-stat grid recomputes the moment you change a slider, so you can see immediately whether your tweak helps or hurts the Sharpe and Max Drawdown. Third, the Send to Visualizer button pushes the recommended basket (with whatever weight tweaks you made) over to the Portfolio Visualizer, where you can swap out individual tickers, change the benchmark, change the analysis period, and re-run a fresh six-stage backtest. Pro and Ultra subscribers can also save scenarios at any step so different weight tweaks live side by side for later comparison.
How does the generator decide whether to include factor tilts?
The Include Factor Tilts checkbox sits in the advanced section of the form and is on by default. When checked, the optimizer can pull factor ETFs into the candidate universe alongside the broad-market core: value (VTV, VLUE), quality (QUAL, SCHQ), momentum (MTUM, IMTM), low-volatility (USMV, SPLV), small-cap value (VBR, AVUV), and similar. The optimizer only actually includes a factor tilt when the historical data shows a positive Sharpe contribution net of fees, so the tilt is data-driven rather than mandatory. When unchecked, the optimizer restricts itself to plain-vanilla broad-market ETFs (VTI, VXUS, BND, VNQ, GLD) and dividend-style ETFs, which produces a simpler “core” portfolio with no smart-beta exposure. Leave it on if you trust the factor literature and want the optimizer to use it where data justifies. Turn it off if you prefer a strict total-market core or have philosophical objections to factor investing.
Why does the target amount feasibility card matter before I generate?
The Goal Feasibility card grades whether your target amount is reachable given your starting capital, your monthly contribution, and the analysis period. It computes the required CAGR (compound annual growth rate) the portfolio would need to deliver to land on the target, then grades it against historical mean-reversion bands. The four grades are Comfortable (required CAGR below 9%, reachable with a balanced mix), Moderate (9 to 12%, requires an equity-heavy mix), Aggressive (12 to 18%, requires a growth tilt and low fees), Very Aggressive (18 to 25%, requires leverage or concentration), and Unreachable (25%+, the math does not support the target). The card matters because the generator will happily produce a portfolio against an unreachable target, but the result will be the highest-volatility variant inside the risk band, which is almost never the right answer. If the feasibility card says Aggressive or Very Aggressive, the honest fix is to raise the monthly DCA, extend the period, or lower the target, not to over-reach inside the optimizer.
How often should I rebalance the portfolio the generator recommends?
For most generated portfolios, quarterly or annual rebalancing back to the target weights is the right cadence. Quarterly is more responsive to drift but increases trading costs and short-term tax events. Annual minimises trading and tax friction but lets the basket drift further between rebalances. Both produce roughly the same long-run return. Two specific cases call for a different cadence. Threshold rebalancing (rebalance whenever any holding drifts more than 5 absolute percentage points from target) is the sharpest option for volatile allocations like Aggressive Growth, because it forces you to trim winners and add to losers exactly when the deviation is largest. Cash-flow rebalancing (direct new monthly contributions toward the most underweight holding) is the cheapest option because it produces zero taxable events and zero trading costs. The Ultra-tier AI Portfolio Rationale card writes a specific rebalancing schedule for each generated basket. The free embed defaults to “quarterly” as a safe baseline.