What Shapes Analysts’ Long-Term Forecasts?

How do professional forecasters think about a firm’s growth far into the future? Marius Guenzel, assistant professor of finance at the Wharton School, discusses his paper “What Drives Very Long-Run Cash Flow Expectations?” which uncovers the different factors that shape the long-term forecasts that drive valuations.

What motivated you to study how professional forecasters form very long-run growth expectations, and what gaps in the existing research did you aim to fill?

Marius Guenzel: One of the main motivations for this research was that while we have a wealth of studies on short-term expectations—how analysts and investors form views on earnings or economic conditions over the next few quarters or years—there’s far less evidence on long-run beliefs. This is surprising because long-term growth expectations are central to firm valuation. More than 70% of a firm’s discounted cash flows typically come from beyond a five-to ten-year horizon, meaning that even small shifts in long-run expectations can have a big impact on valuations.

Another key gap is that short-run and long-run expectations may not be shaped by the same factors. Existing research often assumes that insights from short-term forecasting carry over to longer horizons, but firms’ current performance isn’t always a reliable guide to their long-term trajectory. We wanted to examine whether the determinants of long-run beliefs are fundamentally different—and our findings suggest that they are.

Finally, from a broader perspective, we were also motivated by the idea that long-term expectations are often more subjective and abstract. The further into the future you try to predict, the more individual perspectives, backgrounds, and experiences may shape those beliefs.

Your research finds that long-run growth expectations contain unique economic information. What important data can long-run expectations tell us that aren’t available in short-term forecasts?

Marius Guenzel: Long-run growth expectations capture a different dimension of economic information that short-term forecasts often miss. Short-term forecasts tend to be more reactive, closely tracking recent performance, macroeconomic conditions, and quarterly earnings trends. In contrast, long-run expectations reflect forecasters’ broader views on a firm’s fundamental trajectory—whether it’s positioned for sustained growth, long-term stability, or eventual decline.

One key insight from our research is that long-run expectations are stronger predictors of future earnings growth over extended horizons. We find that while short-term forecasts do a good job of predicting earnings over the next few years, their predictive power fades significantly at longer horizons. By contrast, long-run expectations remain significantly correlated with actual earnings growth even a decade out. This suggests that professional forecasters are capturing meaningful economic signals about firms’ long-run prospects beyond short-term trends.

Another important takeaway is that long-run expectations reflect macroeconomic and structural factors that may not be fully incorporated into short-term forecasts. Forecasters tend to factor in industry shifts, competitive dynamics, and regional growth trends when forming long-term beliefs, making these expectations useful for understanding how market participants think about the broader economic landscape.

Finally, we show that long-run beliefs are shaped by individual forecasters’ perspectives in a way that short-term forecasts are not. Since the future is more uncertain at long horizons, factors like local economic experiences, cultural influences, and personal biases play a larger role. This is important because it highlights that long-term expectations are not just mechanical extensions of short-term trends—they represent a unique and often under-appreciated component of market expectations.

You show that professional forecasters have persistent differences in their long-term expectations. Why do these differences exist, and what do they reveal about our perception of global markets?

Marius Guenzel: A key reason professional forecasters have persistent differences in their long-term growth expectations is that long-run forecasts require more subjective judgment than short-term ones. While near-term projections rely on hard data, long-term expectations depend on qualitative assessments of innovation, industry trends, and macroeconomic shifts, leaving more room for individual interpretation. One major contributing factor is local economic extrapolation—analysts tend to project their own country’s recent economic conditions onto firms, even when those firms operate in entirely different markets. Cultural and institutional differences matter too, with analysts tending to heuristically extrapolate more when evaluating firms they are less familiar with.

These differences have broader implications for global markets—even with widespread access to financial data, forecasters from different backgrounds still interpret the same information differently, suggesting that true financial integration remains limited. Ultimately, as the forecasting horizon extends, expectations become less about precise calculations and more about personal worldviews and narratives, shaping not just firm valuations but also likely affecting how capital moves across markets.

What do your findings mean for investors, companies, and policymakers who rely on long-run forecasts?

Marius Guenzel: For investors, a key takeaway is that long-run expectations contain valuable information but at the same time should be interpreted with caution. Since forecasters systematically differ in their views based on background and geographic exposure, investors should consider whose expectations they are relying on and whether those forecasts may be influenced by factors beyond firm fundamentals. This also means that long-run forecasts may reflect sentiment-driven biases that can create mispricings, especially across regions.

For companies, our results highlight the importance of narrative and positioning in shaping long-term investor expectations. Since long-run beliefs are more subjective, firms can influence their perceived future growth not just through fundamentals but also through strategic communication—emphasizing macro trends, innovation, and expansion narratives that resonate with different investor audiences.

For policymakers, our findings suggest that market expectations about long-term economic growth are not purely data-driven but are shaped by local perspectives. This matters when considering policies that influence investor confidence, global capital flows, or cross-border investments. If long-run expectations systematically vary across regions, capital may not always flow efficiently, and policies aimed at fostering investment might need to consider how global investors perceive long-term growth opportunities rather than just focusing on near-term conditions.

Overall, our research shows that long-run forecasts are not just a reflection of firm fundamentals but also a window into how different forecasters perceive the global economy. Understanding these perspectives can help investors make better decisions, companies shape their messaging, and policymakers design more effective economic strategies.

How can future research build upon your findings?

Marius Guenzel: One key next step is to investigate to what extent the differences in long-run expectations translate into systematic variations in investment decisions—do investors with distinct backgrounds and experiences allocate capital differently based on their long-term beliefs, even when evaluating the same firms? Understanding this could shed light on whether subjective perspectives contribute to persistent mispricings or market segmentation.

Another key avenue is to dig deeper into where investor beliefs actually come from, for example by combining quantitative expectations with the qualitative discussions in analysts’ equity reports. Rather than just inferring beliefs from forecasts, a more direct approach is to elicit economically interpretable narratives directly from text discussions. In follow-up work, we use large language models (LLMs) to do exactly that, extracting structured narratives—capturing topics, valuation channels, sentiment, and time outlook—from thousands of analyst reports. This approach allows us to systematically map how forecasters articulate their reasoning, uncovering which narratives drive forecast disagreement, how shifts in sentiment align with key market indicators, and how optimism shapes the perception of growth versus value stocks. Ultimately, an important next step for research is to bridge the gap between what forecasters believe and why they believe it, offering new insights into the role of subjective narratives in financial decision-making.