Analyze Any Stock or Portfolio with ChatGPT and Claude

Use Copy for AI to turn PortfoliosLab analytics into structured context you can paste into ChatGPT, Claude, or another AI chat.

PortfoliosLab pages are dense with data: returns, risk-adjusted ratios, drawdowns, correlations, capture ratios, and allocation details. Copy for AI turns that data into clean, labeled context you can paste into ChatGPT or Claude, so you can ask what the numbers mean for the assets you hold.

Why context matters for AI

You could type your portfolio holdings into ChatGPT yourself. The problem is that AI models can sound confident even when they are guessing. Paste a screenshot or a few numbers and the model may hallucinate a Sharpe ratio, misread a drawdown, or miss the relevant time period.

Copy for AI gives the model structured, labeled data. Every metric is named, every number is exact, and every period is stated. That makes the answer grounded in your actual figures instead of the model's best guess.

It works on three kinds of page:

To start using the context, you can open a symbol page, for example SMH and copy its entire content as context for AI:

The copied context includes the important numbers from the page in a format that AI can read and interpret.

# VanEck Semiconductor ETF (SMH)

Source: https://portfolioslab.com/symbol/SMH

## Overview

- **Symbol:** SMH
- **Full name:** VanEck Semiconductor ETF
- **Type:** ETF
- **Currency:** USD
- **Market:** US
- **Exchange:** NasdaqGM
- **ISIN:** IE00BMC38736
- **CUSIP:** 92189F676

## Description

The VanEck Semiconductor ETF seeks to replicate the performance of the MVIS US Listed Semiconductor 25 Index, which tracks companies involved in semiconductor production and equipment.

## Price Snapshot

- **Close price:** $627.53

## Classification

- **Issuer:** VanEck
- **Category:** Semiconductors, Technology Equities
- **Asset class:** Equity
- **Region:** Broad
- **Inception:** 2011-12-20
- **Expense ratio:** 0.35%
- **Index tracked:** MVIS US Listed Semiconductor 25 Index
- **Distribution policy:** Distributing
- **AUM:** $67.82B

## Analysis Settings

- **Benchmark:** S&P 500 Total Return
- **Analyzed period start:** 2000-06-06
- **Analyzed period end:** 2026-06-04
- **Risk-free rate:** 0% (applied to all risk-adjusted metrics)

## Returns by Period

| Period         | SMH     | Benchmark |
| -------------- | ------- | --------- |
| 1D             | -1.63%  | 0.42%     |
| YTD            | 74.25%  | 11.36%    |
| 1M             | 20.06%  | 4.61%     |
| 6M             | 74.08%  | 11.27%    |
| 1Y             | 150.04% | 28.58%    |
| 3Y annualized  | 63.96%  | 22.72%    |
| 5Y annualized  | 38.76%  | 14.02%    |
| 10Y annualized | 37.49%  | 15.58%    |

## Risk and Risk-Adjusted Metrics

| Metric                  | SMH     | Benchmark |
| ----------------------- | ------- | --------- |
| Volatility 1M           | 11.58%  | 2.87%     |
| Volatility 6M           | 24.35%  | 9.00%     |
| Volatility 1Y           | 30.57%  | 11.88%    |
| Volatility 5Y           | 35.01%  | 16.90%    |
| Volatility 10Y          | 32.57%  | 18.06%    |
| Sharpe (1Y)             | 4.94    | 2.42      |
| Sharpe (5Y)             | 1.11    | 0.83      |
| Sharpe (10Y)            | 1.15    | 0.87      |
| Sharpe (all time)       | 0.34    | 0.65      |
| Sortino (1Y)            | 5.05    | 3.30      |
| Sortino (5Y)            | 1.66    | 1.25      |
| Sortino (10Y)           | 1.67    | 1.26      |
| Sortino (all time)      | 0.72    | 1.00      |
| Omega (1Y)              | 1.69    | 1.44      |
| Omega (5Y)              | 1.21    | 1.17      |
| Omega (10Y)             | 1.22    | 1.19      |
| Omega (all time)        | 1.09    | 1.14      |
| Calmar (1Y)             | 10.11   | 3.23      |
| Calmar (5Y)             | 0.86    | 0.57      |
| Calmar (10Y)            | 0.83    | 0.46      |
| Calmar (all time)       | 0.14    | 0.21      |
| Martin (1Y)             | 38.76   | 15.09     |
| Martin (5Y)             | 2.40    | 1.67      |
| Martin (10Y)            | 2.96    | 2.18      |
| Martin (all time)       | 0.24    | 0.86      |
| Max drawdown (1Y)       | -14.93% | -8.89%    |
| Max drawdown (3Y)       | -35.74% | -18.75%   |
| Max drawdown (5Y)       | -45.30% | -24.49%   |
| Max drawdown (10Y)      | -45.30% | -33.79%   |
| Max drawdown (all time) | -84.96% | -55.25%   |
| Average drawdown        | -41.08% | -8.16%    |
| Current drawdown        | -1.63%  | -0.32%    |

## Risk / Return Rank

Percentile rank from 0 to 100 (higher is better) comparing risk-adjusted returns against all comparable assets on PortfoliosLab, combining Sharpe, Sortino, and other metrics. The 1Y rank is based on the trailing 12 months.

| Metric              | Rank (0-100) |
| ------------------- | ------------ |
| Combined rank (1Y)  | 96           |
| Combined rank (5Y)  | 93           |
| Combined rank (10Y) | 98           |
| Sharpe rank (1Y)    | 97           |
| Sharpe rank (5Y)    | 95           |
| Sharpe rank (10Y)   | 99           |
| Sortino rank (1Y)   | 95           |
| Sortino rank (5Y)   | 95           |
| Sortino rank (10Y)  | 98           |
| Omega rank (1Y)     | 95           |
| Omega rank (5Y)     | 93           |
| Omega rank (10Y)    | 97           |
| Calmar rank (1Y)    | 97           |
| Calmar rank (5Y)    | 94           |
| Calmar rank (10Y)   | 99           |
| Martin rank (1Y)    | 96           |
| Martin rank (5Y)    | 89           |
| Martin rank (10Y)   | 98           |

## Monthly Return Statistics

| Metric                      | Value               |
| --------------------------- | ------------------- |
| Average daily return        | 0.07%               |
| Average monthly return      | 1.34%               |
| Positive days               | 51.90%              |
| Positive months             | 58.79%              |
| Best day                    | 2025-04-09: 17.16%  |
| Worst day                   | 2020-03-16: -14.41% |
| Best month                  | 2026-04: 32.16%     |
| Worst month                 | 2001-09: -32.51%    |
| Longest monthly win streak  | 7                   |
| Longest monthly lose streak | 6                   |

## Benchmark Metrics

- **Alpha:** 3.55%
- **Beta:** 1.36
- **R squared:** 0.56
- **Upside capture:** 178.10%
- **Downside capture:** 145.77%

## Worst Drawdowns

| Depth  | Peak       | Trough     | End date   |
| ------ | ---------- | ---------- | ---------- |
| 84.96% | 2000-06-22 | 2008-11-20 | 2017-05-09 |
| 45.30% | 2021-12-28 | 2022-10-14 | 2023-07-13 |
| 35.74% | 2024-07-11 | 2025-04-08 | 2025-07-02 |
| 33.62% | 2020-02-20 | 2020-03-20 | 2020-06-05 |
| 27.02% | 2018-03-13 | 2018-12-24 | 2019-04-03 |

## Dividend History

| Period | SMH yield | SMH amount |
| ------ | --------- | ---------- |
| TTM    | 0.2%      | 1.10       |
| 2025   | 0.3%      | 1.10       |
| 2024   | 0.4%      | 1.07       |

## Monthly Dividends

| Period | Jan  | Feb  | Mar  | Apr  | May  | Jun  | Jul  | Aug  | Sep  | Oct  | Nov  | Dec  | Total |
| ------ | ---- | ---- | ---- | ---- | ---- | ---- | ---- | ---- | ---- | ---- | ---- | ---- | ----- |
| 2026   | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 |      |      |      |      |      |      | 0.00  |
| 2025   | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 1.10 | 1.10  |
| 2024   | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 1.07 | 1.07  |

## Sector Exposure

| Sector     | Exposure |
| ---------- | -------- |
| Technology | 100.0%   |

Prompt examples

After copying the context over to an AI chat, you can ask questions about performance, strategy, stress scenarios, or anything else that helps with your analysis.

Ask about metrics or values you don't fully understand

Prompt: "My SMH analysis shows upside capture 178% and downside capture 146%. What does that mean for me?"

Those describe how SMH has moved relative to the S&P 500. When the index rose, SMH rose about 1.8x as much on average; when it fell, it fell about 1.4x as much. Two takeaways. First, both are far above 100%, so this is not a diversifier. It is an amplifier that magnifies the market in both directions. Second, upside capture is higher than downside capture, which is the favorable kind of asymmetry: historically, you were paid more on the way up than you gave back on the way down. But that is a 25-year average. It does not mean any single downturn will be mild. SMH's worst drawdown was about 85%. Holding it is a high-conviction semiconductor bet with a much rougher ride than the index.

Prompt: "Why is SMH's Sharpe ratio 4.94 over one year but only 0.34 all-time?"

Both are real. They just measure different windows. Sharpe ratio measures return per unit of risk, and the last twelve months have been an exceptional stretch for semiconductors: large gains with relatively contained volatility, which pushes the ratio to an unusually high 4.94. A number that high is not sustainable. It reflects a boom, not a normal year. The all-time figure of 0.34 is the steadier view across every market environment since 2000. It is below the S&P 500's 0.65 over the same period, even though the recent period looks much stronger. The lesson is to judge a holding by its long-window ratio, not its best twelve months. The 4.94 is what recency bias looks like in one statistic.

Ask about hidden risk

Prompt: "SMH is up over 150% in the past year. What is the catch I am not seeing?"

The run is genuinely extraordinary, but the trailing return hides the cycle. SMH peaked in mid-2000 and did not reclaim that high until 2017, after a drawdown of roughly 85% that took about 17 years to recover. Its one-year Sharpe ratio looks spectacular at 4.94, yet its all-time Sharpe ratio is 0.34, below the S&P 500's 0.65. Across its full history, you were compensated less per unit of risk than a plain index fund because the booms came with brutal busts. None of that makes it a bad holding. It means today's numbers are the euphoric part of a historically violent cycle. The real question is whether you could sit through a 50-80% decline without selling, because for semiconductors, that has happened before.

Prompt: "How deep have this semiconductor portfolio's drawdowns been, and how long did recovery take?"

Even on a history that starts in 2009, so it never saw the dot-com collapse, this portfolio has had sharp falls. Its worst was about 43%, from late 2021 to October 2022, and it took until November 2023 to fully recover: roughly two years underwater. There was also a roughly 37% fall in 2020 and another in 2011. The pattern is consistent: this is a high-return, high-amplitude holding with beta around 1.4 that periodically gives back a third to a half of its value. If your timeline or temperament cannot absorb a two-year recovery, the headline returns are only half the picture.

Check Diversification

Prompt: "The portfolio has 10 different semiconductor stocks, equally weighted. Is it well diversified?"

Not in the way the count suggests. You hold ten names in equal size, so by headcount it looks spread out. But they all do the same thing, and the data shows it. Every holding correlates with the overall portfolio between about 0.71 and 0.82, and pairs like ADI and TXN move together at 0.82. The portfolio's diversification ratio is only about 1.3 and has been falling over the past decade. In plain terms, ten semiconductor stocks are closer to one large bet across ten tickers than ten independent positions. When the sector falls, they tend to fall together. Real diversification here would mean adding something that does not move with chips: a different sector, bonds, or another asset class, not a different chip name.

Prompt: "Given how correlated these holdings are, are any of them redundant?"

From a diversification standpoint, the most overlapping pair is ADI and Texas Instruments. They correlate at about 0.82, the highest in the book, which makes sense as two analog-focused names. Owning both adds position count without adding much independent exposure. That does not automatically mean selling one; they may differ on fundamentals, valuation, or your own conviction, which the correlation number says nothing about. But if your goal is fewer positions without materially changing the portfolio's risk profile, the closely correlated pairs are where consolidation costs you the least diversification. The bigger point still stands: trimming within semiconductors will not reduce your core risk, which is concentration in a single industry.

Compare ETFs

Prompt: "Is concentrated semiconductor exposure with SMH worth it versus the broader Nasdaq-100 with QQQ?"

SMH has clearly returned more: roughly 37% a year over ten years versus about 22% for QQQ. But you pay for it. SMH's volatility runs about 33% versus QQQ's 22%, and its drawdowns are deeper, with about 45% at worst over the last decade versus 35%. Two things are easy to miss. First, these are not really diversification alternatives. They correlate around 0.85, because QQQ is itself heavily tech, at about 54%, so owning both is doubling down, not spreading out. Second, over their full history since 2000, their risk-adjusted returns are nearly identical, with Sharpe ratios of 0.34 versus 0.41, and both fell more than 80% in the dot-com bust. The real choice is a dial, not a switch: SMH is QQQ's tech bet concentrated further into chips, with more upside in a chip boom and more pain in a chip bust. QQQ also costs less to hold, at 0.18% versus 0.35%.

Explore stress scenarios

Prompt: "Stress-test this semiconductor portfolio against an AI capex slowdown. What would probably break first?"

The first thing to break would probably be the assumption that recent AI-driven demand growth keeps compounding. In a capex slowdown, hyperscalers delay orders, chip inventories build, equipment makers lose visibility, and the whole semiconductor chain reprices at once.

The copied data says this is exactly the kind of portfolio that would feel that shock hard: SMH has beta around 1.36, downside capture around 146%, and historical drawdowns of roughly 45% in the last cycle and 85% in the dot-com collapse.

A realistic stress case is not "the portfolio underperforms by a few percent." It is a fast sector-wide reset where the same names fall together, correlations rise, and risk-adjusted metrics that looked excellent over one year collapse back toward long-run levels. To reduce that risk, model a 30-50% semiconductor drawdown, decide how much loss you can tolerate, then size the semiconductor sleeve accordingly or pair it with assets outside the chip cycle.