Use cases

One tool, many ways to own your conviction

Whether you believe in a theme, want a smoother ride, or have your own way of investing, Indices Lab turns the idea into a portfolio you can test and track. Here are ten ways investors use it, step by step.

The workflow

Six steps, every time

Every use case below runs through the same six steps. Each card highlights the ones doing the heavy lifting.

1Frame

Write the idea in one line and pick what to measure it against.

2Find

Gather companies by ticker, by sector and industry, or from an ETF.

3Weight

Set every weight, or use equal, market-cap or inverse-volatility.

4Test

Backtest the exact portfolio over decades, after costs and tax.

5Stress-check

Correlations, overlap with ETFs, the optimizer, and 10,000 simulated futures with crisis replays.

6Track

Follow it live, get drift alerts, add contributions.

Invest in a theme you believe in

Own the companies behind a trend, your way.

For the believer

01Build a thematic basket

You believe AI, robotics, defense, cybersecurity or the energy transition will grow for years, and you want the companies you think will win, not a fund's list.

Example: Defense: Lockheed Martin, RTX, Northrop Grumman, General Dynamics, L3Harris, equal weights.Illustrative, not a recommendation.

1 Frame2 Find3 Weight4 Test5 Stress-check6 Track
How to build it in Indices Lab
  1. Find: Screen the Aerospace & Defense industry, or start from a defense ETF's holdings and keep the names you want.
  2. Weight: Give pure plays more weight than conglomerates, and watch the sector mix update live.
  3. Test: Backtest against the ETF and the S&P 500, after costs and tax.
  4. Stress-check: Run Overlap against the ETF. 85% the same? You may not need your own index. 40%? You've built a genuinely different view.
  5. Track: Save it and follow it live against the ETF.

Watch out for: Themes get crowded and expensive after the headlines, and a hot theme can still lag the market for years. One theme is one sector: a bad year for it hits everything.

Build your own version →
3-year return+62%S&P 500: +87%
Worst fall−24%S&P 500: −19%

Last 3 years (Oct 2023 to Oct 2026), equal weights, rebalanced quarterly, before costs and tax. Hypothetical.

For the big-picture thinker

02One conviction, the whole value chain

Instead of betting on the one obvious winner, own every link that has to grow if your thesis is right. If AI demand keeps rising, chips, the machines that make them, memory, data-centre power and cloud platforms all need to sell more.

Example: AI: Nvidia (chip design), Applied Materials (equipment), Micron (memory), Vertiv (power and cooling), Microsoft (cloud), one company per link.Illustrative, not a recommendation.

1 Frame2 Find3 Weight4 Test5 Stress-check6 Track
How to build it in Indices Lab
  1. Weight: Give each link of the chain the same share, then split it between the companies in that link.
  2. Test: Compare against holding Nvidia alone and against a semiconductor ETF. What did spreading across the chain cost in return, and how much did it cut the worst fall?
  3. Stress-check: The correlation matrix shows which links really move independently and which are the same bet twice. Attribution shows which link drove the result.
  4. Refine: Run the optimizer with a per-stock cap so no single name takes over.

Watch out for: Every link still depends on the same spending cycle, so a slowdown can hit them all at once. A value chain spreads company risk, not theme risk. The same pattern works for robotics or the energy transition.

Build your own version →
One thesis, five links
NVDAChip design
AMATEquipment
MUMemory
VRTPower & cooling
MSFTCloud
Shared risk: every link depends on the same AI spending cycle
Volatility40.6%Nvidia alone: 46.8%
Worst fall−39%Nvidia alone: −37%

Last 3 years (Oct 2023 to Oct 2026), equal weights, rebalanced quarterly, before costs and tax. Hypothetical.

For the tinkerer

03Rebuild a thematic ETF your way

You like a thematic ETF but not all of it: some holdings are filler, the weights follow a formula you disagree with, and you pay a fee every year.

Example: Start from a cybersecurity ETF, drop the names you don't believe in, and double the weight of the five you do.

1 Frame2 Find3 Weight4 Test5 Stress-check6 Track
How to build it in Indices Lab
  1. Find: Load the ETF's holdings as your starting basket.
  2. Weight: Edit, remove and reweight until it matches your view.
  3. Test: Backtest your version against the original ETF over the same dates.
  4. Stress-check: Overlap shows exactly how far your version has moved from the fund.

Watch out for: You take on the rebalancing the fund did for you, and trading costs replace the fund's fee. The costs and taxes settings show whether it's worth it.

Build your own version →

Control your risk

Smoother rides, steadier income, fewer surprises.

For a steadier ride

04A smoother ride from industries that don't move together

Big drawdowns are what make people sell at the bottom. Industries that respond to different forces, like interest rates, oil, consumer spending and technology cycles, can smooth the ride without giving up on stocks.

Example: Microsoft (technology), Johnson & Johnson (healthcare), Procter & Gamble (staples), NextEra Energy (utilities), ExxonMobil (energy), equal weights.Illustrative, not a recommendation.

1 Frame2 Find3 Weight4 Test5 Stress-check6 Track
How to build it in Indices Lab
  1. Stress-check first: Drop the candidates into a basket and open the correlation matrix. Keep the pairs that move least alike and swap out the ones that duplicate each other.
  2. Weight: Try inverse-volatility weights, which give steadier holdings a bigger share.
  3. Refine: Run the optimizer on lowest volatility and compare it with your own weights.
  4. Test: Compare the worst fall, volatility and rolling returns with the S&P 500. The question isn't "did it beat the market?" but "would I have stayed invested?"

Watch out for: Correlations tend to jump in a crisis, when everything falls together for a while. A smoother ride usually means a lower return in strong bull markets, and past correlations can change.

Build your own version →
How alike they moved, last 3 years
MSFTJNJPGNEEXOMMSFT·-0.16-0.05-0.04-0.04JNJ-0.16·0.380.270.16PG-0.050.38·0.320.06NEE-0.040.270.32·0.19XOM-0.040.160.060.19·
OppositeTogether
Volatility11.8%S&P 500: 15.3%
Worst fall−14%S&P 500: −19%

Last 3 years (Oct 2023 to Oct 2026), equal weights, rebalanced quarterly, before costs and tax. Hypothetical.

For the patient

05A defensive core that pays you

A steady base of established companies with long dividend records, for investors who care more about consistency than excitement.

Example: Coca-Cola, Procter & Gamble, Johnson & Johnson, Duke Energy, Verizon, equal weights.Illustrative, not a recommendation.

1 Frame2 Find3 Weight4 Test5 Stress-check6 Track
How to build it in Indices Lab
  1. Test: Backtests are total return, so reinvested dividends count. Set your dividend tax rate to see what you'd keep.
  2. Compare: Run it against a dividend ETF and the S&P 500, especially through bad years like 2008 and 2022.
  3. Track: Add your monthly contributions and rebalance with new money, so it tops up what's lagging and nothing gets sold.

Watch out for: A high yield can signal trouble rather than value. Dividend stocks often fall when interest rates rise, and a defensive core can lag badly in tech-led rallies.

Build your own version →
Worst fall−12%S&P 500: −19%
3-year return+57%S&P 500: +87%

Last 3 years (Oct 2023 to Oct 2026), equal weights, rebalanced quarterly, before costs and tax. Hypothetical.

For the over-exposed

06Balance out a concentrated position

Many investors hold one stock they can't or won't sell: employer shares, an inheritance, or an early winner that grew huge. The rest of the portfolio should offset it, not double it.

Example: You hold Nvidia through work, 40% of your money. The other 60% goes to Johnson & Johnson, Procter & Gamble, Coca-Cola and ExxonMobil at 15% each.Illustrative, not a recommendation.

1 Frame2 Find3 Weight4 Test5 Stress-check6 Track
How to build it in Indices Lab
  1. Frame: Build one index with the big position at its real weight and the rest around it, so you see the whole picture.
  2. Stress-check: Use correlations to pick holdings that move least with that stock, and check the sector mix so you aren't adding more of the same industry.
  3. Test: Compare the combined portfolio's worst fall with the big stock on its own.
  4. Track: Follow it live, with a drift alert when the big position grows too far past its weight.

Watch out for: If one stock is a large share of your wealth, diversifying around it only helps so much. Tax and lock-up rules on employer shares are personal, so a professional is worth asking.

Build your own version →
Worst fall, combined−17%Nvidia alone: −37%
Volatility, combined18.8%Nvidia alone: 46.8%

Last 3 years (Oct 2023 to Oct 2026), rebalanced quarterly, before costs and tax. Hypothetical.

For the planner

10Live off your own index

You've built or inherited a basket you want to keep, and the question has changed from "how much could it grow?" to "how long will it last if I draw on it?"

Example: $500,000 in a dividend core (Johnson & Johnson, Procter & Gamble, Coca-Cola, PepsiCo, McDonald's, equal weights), taking out $20,000 a year rising with inflation, over 30 years.Illustrative, not a recommendation.

1 Frame2 Find3 Weight4 Test5 Stress-check6 Track
How to build it in Indices Lab
  1. Test: Open the Simulate tab, switch to "Living off it", and enter the pot and the yearly amount.
  2. Stress-check: Read the chance it lasts, the 1-in-10 bad case and the safe yearly amount, and compare them with SPY's on the same page. Replay 2008 on the basket. Switch to Historically to see what the past average alone would say.
  3. Track: Save the index and follow it live.

Watch out for: Stay in Cautious for the decision; Historically assumes the past average continues, which flatters recent winners. A concentrated basket can show a high safe amount and a deep 1-in-10 bad case at the same time, so read both. Nothing here includes tax or your other income.

Build your own version →
Chance it lasts 30 years97%SPY: 95%
Safe yearly amount$24,900SPY: $23,400

Future simulation, Cautious mode, 10,000 futures built from 1993 to 2026, rebalanced quarterly, before costs and tax. Simulated, not a forecast.

Build around your own approach

Your rules, your slice sizes, your values.

For the disciplined

07Core and satellite

Keep most of your money in a broad, cheap core and put a smaller slice into your highest-conviction ideas, so a wrong call can't sink the whole portfolio.

Example: 80% S&P 500 ETF (SPY) as the core, plus Nvidia, Broadcom, Costco and Eli Lilly at 5% each as the satellite.Illustrative, not a recommendation.

1 Frame2 Find3 Weight4 Test5 Stress-check6 Track
How to build it in Indices Lab
  1. Weight: Put the ETF and the stocks in one index with those weights.
  2. Test: Compare with the core on its own. What did the satellite add in return, and in worst fall?
  3. Stress-check: Overlap shows when your satellite picks are already big parts of the core, so you know if you're doubling up.
  4. Track: Drift alerts tell you when the satellite has grown past its slice. Rebalance with your next contribution.

Watch out for: Satellites tend to creep up after a good run. Decide your maximum slice up front.

Build your own version →
80% core, 20% conviction
  • Core: S&P 500 ETF (SPY) 80%
  • Satellite: NVDA, AVGO, COST, LLY at 5% each
3-year return+112%S&P 500: +87%
Worst fall−20%S&P 500: −19%

Last 3 years (Oct 2023 to Oct 2026), rebalanced quarterly, before costs and tax. Hypothetical.

For the principled

08Invest in line with your values

Own the market, minus what you don't want to own: tobacco, weapons, fossil fuels, gambling, alcohol, or any screen that matters to you.

Example: Start from the largest US companies and remove the industries you exclude.

1 Frame2 Find3 Weight4 Test5 Stress-check6 Track
How to build it in Indices Lab
  1. Find: Start from an ETF's holdings or the screener, then remove by sector or industry.
  2. Test: Compare against the unfiltered market to see what your values cost or gained over time. Few investors ever see this number.
  3. Stress-check: The sector mix shows where the weight moved when you removed industries.

Watch out for: You choose what to exclude: Indices Lab doesn't rate companies on ethics or religious compliance. Exclusions can make a portfolio lean heavily into the sectors that are left.

Build your own version →
For the curious

09Prove an idea before you put money in

Read a compelling thesis in a newsletter, a podcast or your own research? Save it as an index, lock today's prices and watch it for a few months before risking a cent.

1 Frame2 Find3 Weight4 Test5 Stress-check6 Track
How to build it in Indices Lab
  1. Test: Backtest it first to see how the idea behaved historically.
  2. Track: Save it so performance counts from today, against the market, with no hindsight.
  3. Decide: After a few months, invest, tweak or drop it. The record stays either way.

Watch out for: A few months is a short time. Use it to learn how the idea behaves, not to judge it.

Build your own version →

For advisors

Model portfolios your clients can read.

For financial advisors

Model portfolios for clients

Build model portfolios, conservative, balanced, growth or around a client's theme, test them with realistic costs and taxes, and hand clients a report with your name on it.

1 Frame2 Find3 Weight4 Test5 Stress-check6 Track
How to build it in Indices Lab
  1. Test: build and backtest each model with your costs and tax rates switched on.
  2. Track: follow the models live and get drift alerts.
  3. Share: export PDF reports branded with your logo and name, or send a share link.
Build your first model →

Example baskets are illustrative, not recommendations. Figures are hypothetical backtests in Indices Lab over the last three years (October 2023 to October 2026), on a total-return basis, rebalanced quarterly, before trading costs and taxes, compared with the S&P 500 (SPY). Past performance does not guarantee future results.

Mix and match

Most investors combine two

Use cases aren't either-or. Keep several indexes side by side, or blend the ideas into one.

Defensive core + AI satelliteA steady dividend base for most of the money, with a value-chain slice for growth.
Values screen + smoother rideRemove what you won't own, then pick the remaining industries that move least alike.
Big position + offsetsKeep the stock you can't sell and build the rest of the portfolio around it.

FAQ

Use case questions

What is thematic investing?

Investing in companies tied to a long-term trend, such as AI, robotics or the energy transition, rather than choosing by country or company size.

Can I build my own ETF?

Not a fund you can sell to others, but you can build your own basket of stocks and ETFs, set every weight, test it and track it like an index. You then buy the holdings through your own broker.

How many stocks should a thematic portfolio have?

Enough that one company can't sink it. Many investors use 10 to 25. The correlation and sector views show whether more names actually add diversification.

How do I find stocks that aren't correlated?

Put the candidates in a basket and open the correlation matrix. Pairs near 0 or below have moved most independently.

Do I need to choose just one use case?

No. Most portfolios combine two or three, and you can keep several indexes side by side.

Does Indices Lab buy the stocks for me?

No. It's a research and tracking tool. You place trades with your own broker.

Are the example baskets recommendations?

No. They illustrate each approach and aren't advice to buy any security. The numbers are hypothetical backtests over the last three years.

Start with the idea you already have

Pick a use case and make it yours. Free to start, no card needed.

Get started free →