# What is a model portfolio? How to read one before you trust it > A model portfolio is a set list of holdings and weights that follows a stated rule and gets tracked over time. Who uses them, the main types, how to read the record (benchmark, backtest vs live, costs, drawdown) and how it differs from an ETF or copy trading. By Maya Koeva · 2026-09-28  A **model portfolio** is a list of investments with a set weight for each, built by following a stated rule and tracked over time as if someone held it. It is a template and a record. You do not own a model portfolio: you read it, compare it with a benchmark, and decide what, if anything, to do with it in your own account. The rule can be as plain as "60% stocks, 40% bonds" or as specific as "the 10 companies whose insiders bought the most stock this month". Either way the rule comes first, the holdings follow from it, and someone keeps score. ## What a model portfolio is Every model portfolio has four parts: - **Holdings.** The stocks, funds or bonds on the list. - **Weights.** How much of the portfolio each holding takes, for example 10% each across 10 stocks. - **A rule.** How the list gets chosen and how often it changes. A good model portfolio states the rule in words you can check against the holdings. - **A record.** How the list did over time, usually next to a benchmark such as the S&P 500. Nothing moves until someone acts on it. An adviser might put a client's money into the model's holdings, a robo-adviser might do it automatically, or you might read the list and research two of the names. The model itself holds no money. ## Who uses model portfolios - **Financial advisers** run model portfolios so that clients with the same goals hold the same mix, and so they can rebalance many accounts at once. - **Robo-advisers** are built on them. You answer questions about your goals and your tolerance for risk, and the service puts you in one of its models, usually made of index funds. - **Newsletters and research services** publish model portfolios as their track record: the stocks they would hold, and how those stocks did. - **Research apps** use them to show what a signal or a data source would have picked, week by week, so you can judge the source by a record instead of by a story. ## Types of model portfolios ### Rule-based or discretionary A **rule-based** (or systematic) model follows a fixed rule every time the list changes. The same inputs produce the same list, whoever runs it. A **discretionary** model depends on a manager's judgment. It can react to things a rule never anticipated, but you cannot test it the same way, because a person's judgment next year may differ from their judgment last year. ### Strategic or tactical A **strategic** model sets a long-term mix, such as 70% stocks and 30% bonds, and returns to it from time to time. A **tactical** model shifts its holdings to follow shorter-term conditions or signals. Tactical models trade more, so costs and taxes matter more. ### Equal weight or cap weight An **equal-weight** model gives every holding the same share: in a 10-stock model, 10% each. A **cap-weighted** model sizes each holding by the company's market value, the way the S&P 500 does, so the largest companies dominate. [Megacaps](/learn/what-is-a-megacap) explains why that matters for an index. Equal weight gives the smaller names on a list the same pull as the giants, which spreads the bets and adds volatility when the smaller names swing harder. ## How to read a model portfolio A return figure on its own tells you little. Check these seven things before you read it. ### What is the benchmark? A model portfolio's return means something only next to what you could have earned for less effort. For US stocks the usual yardstick is the S&P 500, measured with dividends included over the same dates. A stock model compared with a bond index, or with a price index that leaves out dividends, looks better than it is. ### Is it a backtest or a live record? A **backtest** runs the rule over past data to show how it would have done. A **live record** starts on the day the model began publishing its lists in real time, and it cannot be rewritten afterwards. The backtest is useful for understanding a rule. Only the live record shows how the rule behaves when nobody knows the answer yet, because whoever built the backtest already knew how those years turned out. Check where one ends and the other begins, and be wary of a chart that blends them into one line. **[How to verify a track record](/learn/how-to-verify-a-track-record)** walks through the same check for any picker. ### Are costs included? Many published records leave out trading costs, fund fees and taxes. A model that changes its list every week makes dozens of trades a year, and each one carries a spread and possibly a commission. Small, thinly traded stocks cost the most to buy and sell. If a record says "before costs", your own result would come in lower. ### How often does it rebalance, and how much does it turn over? Rebalancing is when the model resets its weights or swaps holdings. Turnover measures how much of the list changes each time. High turnover means more trades, more costs and, in a taxable account, more realized gains. A weekly model and an annual one can hold similar stocks and still produce very different after-cost results. ### How deep was the worst drawdown? Maximum drawdown is the largest drop from a peak to a later low. It tells you what holding the model would have felt like at its worst. A model that finished well ahead of its benchmark but fell 35% along the way only rewarded the people who stayed through the fall. ### Is survivorship bias flattering it? A model tested only on companies that still exist today skips the ones that went bankrupt, got delisted or were bought out, which flatters the result. The same bias applies to the models themselves: providers retire the ones that did badly, so the models you see skew toward the ones that worked. Ask how long a model has been published and whether older versions still show. ### How long is the record? A few months of weekly results can come from a lucky run as easily as from a good rule. Short records also turn into misleading numbers when someone annualizes them: a strong half-year, compounded out to a yearly rate, produces a number the record has not earned. Prefer total return over the stated period, with the number of weeks next to it. Overfitting is the related trap, where a rule gets tuned until its backtest looks great, and [stock screening strategies](/learn/stock-screening-strategies) covers it for screens. ## Model portfolio vs ETF vs copy trading | | Model portfolio | ETF | Copy trading | |---|---|---|---| | What it is | A list of holdings and weights with a record | A fund you buy as one share | Your account mirrors another trader's trades | | Do you own it? | No, you own whatever you choose to buy | Yes, you own fund shares | You own the positions the copying creates | | Who trades | You, your adviser or a robo-adviser | The fund manager, inside the fund | The platform, automatically | | Costs | Your own trading costs, plus any subscription | The fund's expense ratio and trading spreads | Platform fees and spreads, sometimes a share of profits | | What you learn | Which holdings a rule picks, and how they did | The fund's holdings and the index or mandate it follows | What one person did, not why | An ETF packages a set of holdings into one fund you can buy. If a model portfolio is a recipe, an ETF is the dish already cooked. Copy trading hands the decisions to another person's account, so their mistakes land in yours on the same day. A model portfolio leaves the decisions with you, which is slower and gives you the chance to check each name first. ## How to use a model portfolio for research Treat the list as a shortlist. A model portfolio narrows thousands of stocks to a handful that a rule flagged, which is a useful place to start reading. Then: 1. **Read the rule and its caveats.** A model built on insider buys will lean toward smaller companies; one built on Congress filings follows disclosures that can arrive weeks late. [What is insider buying](/learn/what-is-insider-buying) covers where that signal breaks down. 2. **Pick one or two names and research them in full**: the business, the valuation, the bull and bear case. The model tells you a rule chose the stock, and nothing about whether it suits you. 3. **Watch the live record for a while before you trust it.** A few months of weekly results in real time will teach you more about a model than a long backtest. 4. **Size anything you buy for yourself.** A model's weights assume you hold all 10 names. One stock from a list, bought on its own, carries far more risk than a tenth of a diversified list. For most people, a broad index fund belongs at the core of a portfolio, and [how to invest in stocks](/learn/how-to-invest-in-stocks) explains why. Model portfolios fit better as a source of ideas for the part you research yourself. ## Where Quantral fits Quantral has four model portfolios in its Strategies tab. Each is 10 NYSE and Nasdaq stocks in equal weight, rebuilt every Monday before the US open and measured against the S&P 500 with dividends included: - **Insiders**: the companies where insiders spent the most of their own money. - **FinX stars**: what the most accurate voices we track on X are bullish on. - **Capitol**: the companies members of Congress reported buying the most. - **Deep dives**: the companies with the most detailed bullish cases from the voices we follow. Each comes with its current list, the week's changes and a weekly record. The live record started on September 21, 2026, and the earlier hypothetical backtest sits apart from it, labeled. See **[Quantral Strategies](/strategies)** for how the four work. ## Common questions ### What is a model portfolio in simple terms? A list of investments with a set weight for each, chosen by a stated rule and tracked over time. It shows what the rule would hold and how that did, and it holds no money itself. ### Is a model portfolio the same as a fund? No. A fund, such as an ETF or a mutual fund, holds real assets and sells you shares in them. A model portfolio is a list and a record. To own its holdings you would buy them yourself or through an adviser. ### What is the difference between a backtest and a live record? A backtest applies the rule to past data after the fact. A live record starts when the model began publishing its lists in real time and keeps every period since, good and bad. Weigh the live record more, because it could not be tuned with hindsight. ### How often should a model portfolio rebalance? It depends on the rule. Long-term allocation models often rebalance once or a few times a year. Signal-driven models can change weekly. More frequent changes mean more trades, so check whether the record includes costs. ### Should I copy a model portfolio? Copying one outright means taking on its rule, its turnover and its drawdowns without knowing whether they suit you. Using it as a shortlist, and researching the names that interest you, is the safer way to get value from one. ### What is model portfolio investing? Investing by following a model portfolio, usually through an adviser or a robo-adviser that puts your money into the model's holdings and rebalances when the model changes. Some investors follow published models on their own and place the trades themselves. ## The bottom line A model portfolio is a rule written down as a list, with a record attached. Read the benchmark, find where the backtest ends and the live record starts, and check the costs, the turnover, the worst drawdown and the length of the record before the return figure. A model that passes those checks is a good shortlist. What you buy from it is still your decision. --- *This page explains a general investing concept. Quantral surfaces signals and context from public sources to support your own research. Backtested results are hypothetical, and past results do not predict future returns. Nothing here is financial advice or a recommendation to buy or sell.*