Ask AI about your portfolio, and get answers from your real holdings

A portfolio rating is not a verdict on whether your holdings are “good”. It is a summary of measurable properties: concentration, sector and currency spread, income reliability, cost, volatility relative to a benchmark, and how far the portfolio has drifted from the target allocation you set.

Strength and weakness are only meaningful against a stated objective, so Ask asks for one: a capital target, an income target, or a date. It then reports which measured properties help that objective and which work against it, without telling you what to hold.

A deceptively simple question with four possible meanings: today, this year, since you started, or versus an alternative. Ask answers all four, because a portfolio can be up on the day and behind its benchmark over a decade at the same time.

“Weakest” is resolved by ranking measurable gaps rather than by opinion: largest single-name weight, largest share of income from one payer, biggest drift from target allocation, worst risk-adjusted return, and the holding with the poorest contribution to the stated goal.

Ask deliberately does not issue a single score. A one-number rating hides the trade-off that actually matters: a portfolio optimised for income scores badly on growth measures, and the reverse. It also reads like a recommendation, which this is not.

Phrased as a joke, answered as an audit. Ask returns the same measured observations it would give a polite question: concentration, drift, income dependence, cost drag, benchmark gap. The verdict language stays absent, because entertainment framing does not change what the numbers say.

Snowball Analytics hosts public portfolios you can open and read in full: holdings, weights, yields and returns. Ask can contrast your measured properties against a named public portfolio, which is a far more concrete comparison than an abstract benchmark.

The questions that pay off are the ones a spreadsheet cannot answer in one step: where income actually comes from, what happens to the plan if a contribution changes, how much of a return came from one position, and how far reality has drifted from the target.

Ask will not label a portfolio suitable or unsuitable for anyone, because that is a personal-recommendation question that depends on circumstances no tool can see. What it can do is describe complexity: number of positions, number of currencies, how much sits in single names versus funds.

Ask can compute exactly what changes if a position is removed: new weights, new yield, new income total, new sector split, shown before and after side by side. It stops there, because whether that constitutes an improvement is a judgement about your objectives.

Forward income is the sum of each holding's declared or expected distribution multiplied by the shares you actually hold, adjusted for known increases and for the payment calendar. It is not the trailing figure, and it is not the yield multiplied by portfolio value.

Three yields exist and they answer different questions. Current yield divides forward income by today's value. Pre-tax yield ignores withholding. Yield on cost divides the same income by what you originally paid, so it rises over time even if nothing changes.

Income concentration and capital concentration are different measures, and the gap between them is usually the more interesting number. A high-yield position can be a small slice of capital while supplying most of the cash.

Monthly income is rarely level. Ask returns the forward calendar month by month, which exposes the quarterly clustering most portfolios inherit from US payers and the annual lump many European holdings pay in a single month.

Received income is read from your transaction history rather than modelled: actual cash, on actual dates, in the currency it arrived in. This is the figure that reconciles with a broker statement; the forward figure will not.

The answer needs four inputs: today's income, your contribution rate, an assumed dividend growth rate, and whether distributions are reinvested. Ask asks for the ones you have not supplied and shows the date each assumption produces, rather than a single confident number.

Extra monthly income divided by the yield of what you buy gives the capital required. The sensitivity to yield is the part most people underestimate: the same $10 a month costs four times as much at a 1% yield as at 4%.

Income growth has two sources that are worth separating: organic growth, where companies raise their distributions, and contributed growth, where you buy more shares. A portfolio can show rising income while every holding freezes its dividend.

Yield on cost measures today's income against the price you originally paid. It rises automatically whenever a holding appreciates or raises its distribution, and it never falls when prices rise, which is why it flatters portfolios held a long time.

Sustainability is assessed from payout ratios, distribution history, and how much of total income depends on a single payer or a single strategy. Ask reports the dependency; it does not forecast a cut or advise a change.

There is no single answer, so Ask returns three: a bearish, neutral and bullish path built from a return assumption you can see and change. The spread between them is the honest content of the answer, because a single number would imply precision that does not exist.

One projection, three assumptions. Each scenario uses a different compound return, holds contributions constant, and is drawn as a path rather than a single endpoint, so you can see the point at which the three diverge.

Contributions and compounding pull in different directions over different horizons: new money dominates the first few years, compounding dominates the last. Ask shows the two contributions to the final figure separately.

Doubling time depends only on the compound rate, not on the starting amount, which is why the answer is the same for a $5,000 and a $500,000 portfolio. Contributions shorten it, and Ask reports both versions.

A Monte Carlo run replaces a single assumed return with thousands of randomly sequenced ones, and reports the distribution of outcomes instead of a line. Its value is that it exposes sequence risk, because the same average return can end very differently depending on the order.

Ask handles recurring contributions in any supported currency, at any interval, and can apply them from a future start date. The result is shown as a path with the contributed total marked separately from growth.

Only if you say so, and Ask states which mode it used. Price-only, dividends-taken-as-cash and full reinvestment produce materially different paths, and conflating them is the most common error in DIY projections.

This is the projection run backwards: fix the target and the date, solve for the contribution. Ask returns the monthly figure for each return assumption, so you can see how much of the plan depends on markets rather than on you.

Long horizons magnify the assumption rather than the starting balance, so Ask widens the reported range rather than narrowing it. At 30 years the difference between a 5% and a 9% assumption is larger than the entire starting portfolio.

Probability language requires a model and a set of assumptions, and Ask shows both. It reports how a target is affected by contribution size, time and assumed return, and it is explicit that a probability is a property of the model, not a prediction about markets.

On track is measured against a goal you define: a capital figure, an income figure, or a date. Ask compares the required compound rate implied by your goal with the rate your portfolio and contributions currently produce, and reports the gap in both directions.

The coverage question needs a monthly expense figure from you. Ask then reports the crossover month under your contribution and dividend-growth assumptions, and shows how sensitive that date is to each input separately.

Required capital is the annual figure divided by the withdrawal or yield rate you assume, and the assumption dominates the answer. Ask shows the figure across a range of rates rather than picking one for you.

This is the required-rate calculation, and it is the most useful reality check available. Ask solves for the compound rate your goal implies, then places that rate next to what the portfolio has actually delivered.

Ask recalculates the income and growth paths for a hypothetical weight change and shows both, because the trade is almost always income now against capital later. It presents the two paths; it does not say which to prefer.

Enough is defined only by a goal and a date. Ask converts the goal into a required monthly contribution at several return assumptions, and compares it with what you actually contribute.

A FIRE number is annual spending multiplied by the inverse of a chosen withdrawal rate. Ask computes it from your figure, shows current progress as a percentage, and reports the implied date under your contribution rate, flagging that the withdrawal rate is an assumption, not a law.

Ask will show the arithmetic: projected value at your retirement date, sustainable withdrawal at several rates, and how long the capital lasts at your spending level. It will not answer yes or no, because that is a personal recommendation.

Withdrawals reverse compounding, and the damage depends heavily on when they start. Ask models a fixed annual or monthly withdrawal against the same return assumptions and reports the year the balance peaks and the year it depletes, if it does.

Ask reruns the plan with a contribution pause and reports the delay in months rather than the change in end value, because a date is what most goal questions are really about.

A fair comparison needs the same measure on both sides. Ask uses time-weighted return, which strips out the effect of your deposits and withdrawals, and shows the money-weighted figure separately because that is what your account actually earned.

Time-weighted return measures the holdings; money-weighted return (IRR) measures your decisions, including when you put money in. A well-timed deposit before a rally raises IRR without any holding performing better.

Attribution answers this by decomposing the gap into its sources: which positions dragged, which helped, how much cash sat idle, and how much came from currency. Ask reports the decomposition rather than diagnosing a cause.

Ask ranks holdings by contribution to portfolio return, not by percentage change, since a 200% gain on a 0.1% position moves almost nothing. Both views are shown, because they answer different questions.

Ask separates return into three sources: price movement on holdings you kept, realised gains on positions you closed, and income received, so that “what made the difference” has a numerical answer rather than a narrative one.

Yes: an index, an ETF, or a portfolio of your own construction. Comparing against a portfolio that matches your actual asset mix is usually more informative than comparing a 24% cash-and-gold portfolio to a pure equity index.

Drawdown from peak is measured on portfolio value adjusted for cash flows, so a deposit does not disguise a decline and a withdrawal does not manufacture one. Ask reports current drawdown, the deepest historical drawdown and how long recovery took.

Total return splits cleanly into income and capital appreciation, and the split varies enormously between portfolios. Ask reports both, in currency and as a share of the total.

IRR is the annualised rate that makes all your cash flows balance against today's value, the closest thing to “what my money actually earned”. It is sensitive to contribution timing, which is a feature rather than a flaw.

Almost always a definition difference rather than an error: brokers usually report simple return on current holdings, frequently exclude dividends already withdrawn, and rarely use time-weighting. Ask can show each definition side by side so the discrepancy resolves.