ECOPETROL (2010–2025): FROM VALUE DESTRUCTION TO YIELD OPPORTUNITY — VOLATILITY, CYCLICITY, AND THE POWER OF A RECORD DIVIDEND.

Ecopetrol has been a structurally volatile stock with weak returns in the very long term, despite periods of strong gains linked to oil cycles and high dividends. The data you provide for 15 years shows a cumulative negative return of approximately -23.25% and an average monthly return of -0.17%, confirming that, for a buy-and-hold investor, the stock has destroyed value in price terms, even before adjusting for inflation.

The historical average price is around COP 2,782, with a median of 2,553 and a mode of 1,395. The maximum was 5,850 and the minimum 881, with a range of 4,969 COP and a very high standard deviation (≈1,111), indicating high structural volatility.

This trend aligns with internationally observed behavior: Ecopetrol went from peaks linked to the commodities supercycle (2010–2013) to sharp declines following the oil price collapse of 2014–2016, the 2020 pandemic, and subsequent episodes of political uncertainty in Colombia.

The positive skewness (0.70) and kurtosis close to 0 indicate a distribution with moderate tails; it has experienced more episodes of sharp rises than extreme falls, but the declines have been persistent enough to leave the 15-year balance in negative territory.

The underlying reasons for this diminishing return are the near-total dependence on international oil prices and the peso-dollar exchange rate; downward shocks to crude oil erode profits and investment capacity.

High CAPEX requirements for exploration and refining limit free cash flow growth during bear markets.

As a state-owned company, the market anticipates potential increases in the tax burden, changes in dividend policy, or pressure to finance public programs, which raises the risk premium demanded by investors.

In the table, the 90-, 180-, and 360-day price forecasts fall from COP 1,726.53 to COP 1,586.33, implying a negative slope over one year and confirming that the model projects a bearish bias.

The linear correlation of 52.44% suggests that the linear trend explains only half of the price variation; the rest is noise and regime changes, typical of cyclical assets. The sixth-order polynomial correlation (85.48%) indicates that the best-fit curve is "waved": periods of recovery followed by further declines, without establishing a sustainable upward trend.

At a recent fundamental level, Ecopetrol has improved its results thanks to periods of high oil prices and the acquisition of ISA, but it has also increased its debt and exposure to energy transition risks, which means the market does not consistently reward it with high multiples.

In other words, over several years, the stock tends to rebound when crude oil prices rise and decline when they fall, without consolidating a long-term upward channel.

In the short term, the 95% confidence level (≈34.9 COP) compared to the volatility of 1,110 COP shows that the prediction intervals are relatively narrow only when working with averages, but the daily price can deviate significantly from the expected path.

 The fact that the expected value X=1,915 COP is clearly below the historical average of 2,782 COP, along with p=21.76% and q=78.24%, can be interpreted as indicating that, under the probabilistic model, there is a relatively low probability that the price will consistently exceed the historical average in the short term.

The decreasing 90–180–360 day forecasts reflect a bearish continuation bias: the model "reads" that the later segments of the series have a negative slope and projects them forward, consistent with a context of lower growth expectations for the fossil fuel sector and greater environmental regulatory pressure.

In the short term, therefore, behavior is dominated by news about Brent and WTI crude oil prices. Domestic political signals (tax reform, oil policy, discussions on exploration and energy transition) also play a role. Dividend announcements, which usually trigger temporary spikes, do not reverse the underlying trend.

Why does the performance appear “always declining”? Connecting the data with the asset's history, the “feeling” of permanent decline stems from a combination of factors, including a high starting point. Those who evaluated the asset 15 years ago likely used areas near the peaks of the oil supercycle as a reference point; since then, even recent rebounds have not fully recovered those levels Unlike companies such as large global technology firms, whose value grows with the expansion of digital markets, Ecopetrol depends on a commodity (oil) whose global consumption is likely to stagnate or decline due to the energy transition, limiting the "growth premium" the market is willing to pay.

 

The higher risk premium implies lower valuation multiples and, therefore, less dynamic price trajectories, even when one-off earnings are strong.

The large standard deviation and polynomial correlations indicate that the price fluctuates significantly, but around an axis that, rather than rising, tends to slope downward over the measured time horizon.

For a long-term investor, Ecopetrol behaves more like a yield equity asset (high dividends in favorable cycles) than a growth stock; the main source of return is not capital appreciation but periodic dividends, which have been very high in years of record earnings.

For the medium and short term, the appeal lies in trading cycles: buying during panic periods when oil prices and sentiment regarding Colombia are depressed, and selling during euphoric phases. The model itself, with its downward-trending forecasts, suggests caution now and waiting for better entry points.

The key risk going forward is the energy transition and the regulatory framework. If the company manages to diversify into low-carbon energy and maintain financial discipline, it could stabilize its trajectory; otherwise, the baseline scenario will remain one of high volatility with a weak real trend.

Ecopetrol's stock has shown declining long-term performance due to a combination of high dependence on the oil cycle, political risk, and a lack of a clear structural growth narrative. Statistics and forecasts confirm that even during rallies, the model continues to show a bearish bias, so the stock should be treated more as a cyclical dividend asset than as a bet for sustained growth.

The first chart shows the evolution of Ecopetrol's (EC) stock price from the start of the study on January 4, 2010, with a share price of 2,495, to the end on December 4, 2025.

The stock exhibits a general long-term downward trend, despite notable fluctuations, such as the initial peak and the subsequent rebound that fails to reach the starting price. The difference between the initial and final prices indicates a net loss in the stock's value during the analyzed period. The second graph confirms this trend with a calculated return of -23.25% for the period, a crucial figure that demonstrates the asset's poor performance.

The linear trend, or first-order regression, shown in red, indicates a downward overall direction. The coefficient of determination (R² = 0.273) is low, indicating that time explains only 27.3% of the price variability. This suggests that other non-linear factors (economic cycles, oil prices, politics) are much more important in determining the price.

The polynomial trend lines (third and sixth order) fit the price behavior much better. The fact that the polynomial correlations are high (65.48% for the third order and 92.70% for the sixth order) is a key finding. This implies that the stock does not follow a simple or random path, but rather moves in complex cycles or patterns (like waves) that are effectively captured by higher-order models. The 6th-order polynomial model has excellent predictive and descriptive power regarding the stock's cyclical pattern.

The descriptive statistics table provides an in-depth analysis of Ecopetrol's historical price distribution. The high standard deviation relative to the mean confirms that Ecopetrol is a volatile and high-risk asset.

These two measures, skewness and kurtosis, are fundamental to understanding the shape of the price distribution, contrasting it with the normal distribution (Gaussian bell curve). Positive skewness means that the tail of the distribution is more extended to the right (higher values). In terms of stock prices, this implies that there are more extreme prices above the mean than below, but the greatest concentration of data is found at the lower values ​​(near the mode). The Gaussian bell curve, being symmetrical, has a skewness of zero, so Ecopetrol's distribution is not normal.

Negative kurtosis indicates a platykurtic distribution (flatter than normal), meaning the distribution has lighter tails (fewer extreme values ​​at the peaks) and is more dispersed around the mean compared to a normal distribution.

Studying kurtosis is critical for risk management (Value at Risk - VaR), since most financial models assume a normal distribution.

Positive kurtosis (leptokurtic) implies heavy tails, meaning a higher probability of extreme events (large losses or gains) than a normal distribution would predict. This increases risk.

Negative kurtosis, or platykurtic, implies light tails, suggesting that extreme events are less likely compared to a normal distribution. This could be perceived as a lower risk of sudden and massive price shocks. However, negative kurtosis also means the data is more spread out, which aligns with the high standard deviation.

 

The histogram and frequency polygon corroborate the findings regarding skewness and kurtosis. The polygon curve is clearly skewed and does not conform to the bell shape of a normal distribution. The peak or mode is found in the lower price ranges (881-1880 and 1881-2880), confirming the positive skew: most of the time, the price was at the lower end of its range.

The tail extends progressively to the right, indicating higher values, which is evident in the lower number of accounts in the upper ranges (3881-4880 and 4881-5880).

Ecopetrol's price series does not follow a normal distribution. It is positively skewed and platykurtic, meaning that risk cannot be accurately modeled assuming normality. The high correlation of the polynomial models suggests that investment should be based more on the analysis of cycles and cyclical trends than on a random walk model.

Based solely on this first-stage statistical and econometric analysis, a negative historical return is found, and its long-term return of -23.25% is a sign of poor performance. High volatility and a large standard deviation indicate that it is a high-risk asset with significant price fluctuations.

The success of the polynomial model (92.70% correlation) suggests that the stock is cyclical. If the current price (close to 1915) is at a trough in a cycle (and the mode is at 1385.00), it could be a good entry point while awaiting the next cyclical rebound. The 1915 price is significantly lower than the mean (2781.60), suggesting that the stock may be historically undervalued. Strictly from a statistical and econometric perspective of this first stage, the investment presents a high risk due to volatility and negative historical returns. However, the high polynomial correlation and the current price suggest that the stock may be at a low point in a cycle. To make an informed decision, it is absolutely necessary to complement this analysis with a fundamental analysis of the industry and future oil prices (Brent/WTI), Colombian energy policy, and Ecopetrol's financial results.

Technical analysis, which examines the current price in relation to important moving averages and support/resistance levels, is not included here.

Technical analysis (indicators and moving averages) is used to assess market sentiment, trend direction, and price momentum.

Moving averages are indicators that smooth price action to identify trend direction. The fact that all 12 moving averages (Simple and Exponential) for all periods (5, 10, 20, 50, 100, and 200) show a "Buy" signal is an extremely bullish conclusion.

The current stock price (recent close 1,940) is above its short-, medium-, and long-term historical average prices. This confirms that Ecopetrol is in a well-defined and sustained upward trend, overcoming the statistical concern of its very long-term negative historical returns.

Oscillators measure the speed and change in price movements (momentum). The summary of 9 Buys, 1 Neutral, and 1 Sell is predominantly bullish, but introduces an element of caution.

The trend is strong buying (driven by the moving averages), but there is a risk of a short-term correction (given by the selling ROC and the near-overbought RSI). The current price (1,940) is located right around the classic pivot point (1,910), implying that it is in an equilibrium zone where support (S1: 1,900) and resistance (R1: 1,915) are very close.

The cumulative probability (p = 21.75%), if we interpret this data as the probability that the price has historically been at or below current levels (the "left tail" of the distribution), confirms what the statistical analysis suggests: the current price (1,940) is significantly below the historical average (2,781.60). The fact that it is in the lower quartile of its historical range (21.75%) is a sign that the stock could be undervalued and in a low point in the cycle (considered a "good stock" to buy).

The beta of 0.17 is a critical risk management metric; a beta of 0.17 is extremely low, meaning Ecopetrol's stock is highly defensive and has very little correlation with the movement of the overall stock market (indices like the S&P 500 or MSCI). Despite the high historical standard deviation we identified, the systematic risk (beta) is very low.

The fundamentals present the strongest evidence in favor of the investment; the 19.4% dividend yield is exceptional and offers income returns that significantly offset any capital fluctuation risk. For income investors, this is extremely attractive.

The price-to-earnings ratio (P/E) of 6.9x is very low, suggesting that the stock is undervalued relative to its earnings per share (EPS of 277.25). The stock is cheap. The price-to-book ratio of 1.0x indicates that the stock is trading at virtually its book value, another indicator that it is not overvalued.

The 22.7% year-over-year change confirms that the company has experienced a strong recent rally, reversing the long-term historical trend. Investing with a medium- to long-term horizon is essential, and the dividend yield provides a safety net and a source of income that is hard to match.

Technical and statistical analysis suggest that the price is near a short-term resistance/overbought level. If you wish to mitigate the risk of a small pullback (due to the RSI), consider this (high), one could consider entering in stages or waiting for a small pullback towards the 1,900 support level (S1) or the 5-month moving average (MA5) (1,905), although the risk of missing the upward momentum is real. Investment risk (volatility) is offset by defensive stability (low beta) and income profitability (high dividend).

The value factors and dividend yield are so strong that they outweigh concerns about historical volatility and potential short-term technical overbought conditions. It is a very favorable time to consider investing in Ecopetrol.