Sell/Buy Ratio
In this study, we analyze a widely used metric featured in our Basket Analysis Tool — the sell/buy ratio — which helps quantify insider sentiment by comparing the total value of shares sold to those purchased within a given period.
Why Exactly Sell/Buy?
Highlights Selling Pressure
A higher sell/buy ratio directly signals more selling than buying, which is often interpreted as bearish sentiment.
Example: A ratio of 3 means insiders are selling 3x more than buying → clear negative signal.
Magnifies Risk
It aligns with risk detection: investors want to spot unusual selling activity compared to buying.
Since buying is often more selective and confident (especially for insiders), comparing selling to buying can reveal deterioration in sentiment faster.
Consistent With Other Financial Ratios
It mirrors other risk-oriented metrics: like debt/equity or put/call ratio — where a higher ratio implies higher caution or risk.
What Is a Normal Value for the Sell/Buy Ratio?
In this post, we introduced the law of large numbers as a foundation for calculating the UBSTA value.
In reality, each instrument has its own distribution shape and corresponding Sell/Buy (S/B) ratio values across the first and second standard deviations. Even an ETF and a stock contained within it can exhibit completely different S/B ratio distributions.
Global Distribution of S/B ratio
The period of the last 36 months was selected intentionally.
Be careful when looking at the current month's values, because the data is continuously being populated, and you may see extremes simply due to incomplete data.
To put this S/B ratio chart into perspective, let's compare it with a broad market instrument - VT, one of the largest and most popular ETFs globally.
It doesn't take a genius to realize there's no clear evidence of a strong correlation between index or ETF performance and the S/B ratio. Indexes tend to rise over time for a variety of reasons unrelated to that metric.
One of the simplest and most straightforward explanations for why extreme values in the S/B ratio often coincide with rising performance of a given instrument is that insiders may sell their positions when the price is simply too good to resist.
The complexity of interpreting insider activity is precisely one of the reasons why the SIT project was created. That's why we recommend focusing on individual stocks where insider activity can offer clearer signals.
Let's take a look from another angle. VT is an ideal instrument for this analysis because the S/B ratio shown above represents all the countries supported by the SIT platform. We analyze VT's performance following each monthly S/B ratio reading, focusing primarily on 1-, 3-, and 6-month periods — similar to how our ML model 'StockSense' operates.
Study Results
- Mean: 13.4116
- +1 Std Dev: 61.6432
- +2 Std Dev: 109.8748
| Month | S/B | 1M | 3M | 6M |
|---|---|---|---|---|
| 06.2022 | 3.2403 | 6.73% | -7.14% | 2.23% |
| 07.2022 | 1.1065 | -4.05% | -7.67% | 1.68% |
| 08.2022 | 1.0056 | -9.54% | 1.44% | 4.24% |
| 09.2022 | 0.2206 | 6.37% | 10.09% | 18% |
| 10.2022 | 1.1173 | 4.67% | 9.34% | 11.58% |
| 11.2022 | 0.9335 | -1.82% | 2.76% | 5.43% |
| 12.2022 | 1.9366 | 6.41% | 7.19% | 13.64% |
| 01.2023 | 1.4894 | -1.63% | 2.04% | 10.77% |
| 02.2023 | 0.4843 | 2.4% | 2.6% | 9.41% |
| 03.2023 | 0.3294 | 1.3% | 6.02% | 1.74% |
| 04.2023 | 1.0729 | -1.22% | 8.42% | -2.57% |
| 05.2023 | 1.2926 | 5.82% | 6.64% | 7.73% |
| 06.2023 | 0.9387 | 3.72% | -4.04% | 6.46% |
| 07.2023 | 0.1452 | -2.84% | -10.13% | 4.8% |
| 08.2023 | 0.2243 | -4.77% | 1.03% | 10.88% |
| 09.2023 | 0.9625 | -3.4% | 10.34% | 19.67% |
| 10.2023 | 2.7734 | 9.23% | 16.62% | 19.76% |
| 11.2023 | 12.993 | 4.57% | 9.75% | 14.69% |
| 12.2023 | 18.5416 | 1.23% | 7.54% | 10.61% |
| 01.2024 | 0.0008 | 2.79% | 2.69% | 11.28% |
| 02.2024 | 28.6655 | 3.35% | 4.5% | 8.6% |
| 03.2024 | 0 | -3.58% | 2.59% | 9.27% |
| 04.2024 | 12.4957 | 4.6% | 8.36% | 12.49% |
| 05.2024 | 11.4373 | 1.72% | 3.92% | 10.37% |
| 06.2024 | 10.2492 | 1.98% | 6.66% | 5.7% |
| 07.2024 | 16.4794 | 0.32% | 3.81% | 7.33% |
| 08.2024 | 0.0003 | 2.2% | 4.12% | 2.8% |
| 09.2024 | 0.0006 | -0.74% | -0.9% | -2.09% |
| 10.2024 | 7.0301 | 2.64% | 3.39% | -0.81% |
| 11.2024 | 1.7137 | -2.72% | -1.26% | 2.93% |
| 12.2024 | 0.0076 | 3.55% | -1.21% | 9.5% |
| 01.2025 | 4.2482 | -1.98% | -4.06% | 6.9% |
| 02.2025 | 3.1839 | -2.67% | 4.25% | 11.48% |
| 03.2025 | 11.5785 | 0.56% | 10.84% | 18.84% |
| 04.2025 | 2.1528 | 6.51% | 11.43% | % |
| 05.2025 | 10.6332 | 4.17% | 7.65% | % |
| 06.2025 | 5.2042 | 1.1% | 7.22% | % |
| 07.2025 | 0 | 2.22% | % | % |
| 08.2025 | 8.9302 | 3.01% | % | % |
| 09.2025 | 35.4344 | % | % | % |
| 10.2025 | 302.7996 | % | % | % |
| Month | S/B | 1M | 3M | 6M |
Pearson Correlation
As a final step, let's calculate the linear correlation (Pearson correlation coefficient) between the S/B ratio and the forward performances over 1 month, 3 months, and 6 months.
S/B vs 1M Perf: 0.1904
S/B vs 3M Perf: 0.2481
S/B vs 6M Perf: 0.1663
Looking at the results, the main observation is that the correlation between the S/B ratio and future performance is positive, indicating a relationship in the expected direction, but it is not particularly strong.
Knowledge Base
Mean (Average): The mean represents the central value of the data. It is the sum of all observed values (in this case, the Sell/Buy ratios) divided by the number of data points. The mean gives us an overall sense of the "typical" value in the dataset. For example, a mean value of 4.24 indicates that the average Sell/Buy ratio over the selected period is around 4.24.
Standard Deviation (std): Standard deviation measures the spread of the data points from the mean. A larger standard deviation indicates that the data points are more spread out, while a smaller standard deviation suggests they are clustered around the mean. For example, a standard deviation of 6.16 means that most of the Sell/Buy ratios fall within 6.16 units of the mean, but there can be significant variation as well.
Standard Deviation 1 (std1): These values represent the range within which approximately 68% of the data points should fall if the distribution is normal.
Standard Deviation 2 (std2): The second standard deviation (mean ± 2 * std) extends the range, covering approximately 95% of the data points in a normal distribution.
Takeaways
1. S/B ratio interpretations, especially those circulating on social media, should be evaluated critically and with practical reasoning.
2. Stock markets really do grow most of the time. It's very difficult to go against the broader market trend.
3. This study is based on historical data, which means that unexpected events—so-called “Black Swans”—cannot be anticipated in advance.
4. Trust your observations — a Sell/Buy ratio greater than 1 can indeed represent the mean.
Data updated weekly