The core insight behind FTSformer is deceptively simple but technically demanding. Most existing forecasting architectures, including sophisticated transformer models that have dominated long-horizon time series benchmarks, concentrate almost exclusively on the historical behavior of the target variable. They learn the internal dynamics of a single asset and extrapolate from them. Yet financial time series are shaped by an intricate web of external drivers: macroeconomic indicators, sector indices, and cross-asset signals all exert latent coupling effects on the target. When these exogenous variables are ignored or treated as generic side inputs, models lose critical information about why a market moves, not merely how it moved. The researchers set out to build an architecture that simultaneously captures cross-asset interactions, multi-scale temporal dynamics, and the notoriously complex loss landscape that deep forecasting models must navigate.

At the heart of the framework lies the Dual-Domain Fusion Layer, or DDF, a module designed to extract and fuse heterogeneous exogenous information in both the time domain and the frequency domain. Time-domain features capture raw temporal patterns such as trends and sudden shocks, while frequency-domain representations expose periodicities and oscillatory components that are invisible in raw waveforms. By jointly processing both domains and explicitly emphasizing their intrinsic heterogeneity, the DDF layer allows the model to distinguish, for example, between a slow-moving macroeconomic drift and a fast, cyclical sector rotation that may both influence the same stock. This dual-domain strategy reflects a broader trend in modern time series research, where frequency-enhanced architectures such as FEDformer have shown that spectral information can substantially improve long-term forecasting, but extends it by systematically integrating external signals rather than only the target series.

Complementing the fusion layer, the researchers introduce a Multi-Scale Temporal Partition mechanism, abbreviated MSTP, together with a Channel-Modulation Preprocessor known as CMP. Financial data unfold at multiple characteristic timescales simultaneously: intraday noise, weekly rhythms, monthly cycles, and secular trends all coexist in a single price series. The MSTP mechanism partitions the input sequence across these scales so that the model can attend to short-lived fluctuations and long-range dependencies without forcing one convolutional or attention window to serve all purposes. The CMP, in turn, addresses a practical problem that plagues multivariate forecasting: different input channels arrive with vastly different signal strengths. A dominant channel can overwhelm weaker but informative ones during training, destabilizing gradients and slowing convergence. The CMP dynamically adapts the weights assigned to each input channel, balancing signal strengths before the core temporal modeling module ever sees the data. According to the authors, this preprocessing step measurably enhances both training stability and convergence speed, an advantage that becomes evident in the ablation experiments reported in the paper.

The temporal backbone of FTSformer is a component the authors call BA-LSTM, a bidirectional long short-term memory architecture tasked with capturing both short-term market fluctuations and long-term dependencies. LSTMs, first popularized in the deep learning era for sequence modeling, remain a powerful tool for financial prediction, and hybrid designs combining recurrent networks with transformers have proliferated in recent literature. In FTSformer, the BA-LSTM operates in concert with the transformer’s attention machinery, providing a recurrent pathway that preserves order-sensitive information while the attention layers model global interactions across the multi-scale, exogenously enriched representations. This division of labor echoes a recurring theme in the forecasting literature, where purely attention-based models sometimes struggle with the strong nonlinearity and regime changes characteristic of financial data.

The empirical evaluation is unusually broad for a methods paper in this domain. The team tested FTSformer on multiple public benchmark datasets as well as a real-world Canadian traffic dataset, deliberately stepping outside finance to probe cross-domain transferability. The headline result is a reduction in forecasting errors of up to 13.4 percent on the Exchange dataset, a benchmark notorious for its nonstationary exchange-rate dynamics that punish rigid models. Across the full suite of public benchmarks, the model achieved average reductions of 1.54 percent in root mean square error and 2.22 percent in mean absolute error over the strongest baseline. While single-digit average gains may appear modest, in competitive forecasting benchmarks where state-of-the-art models are separated by fractions of a percent, such improvements are meaningful, and the consistency of gains across heterogeneous datasets suggests the architecture captures something genuinely general rather than dataset-specific.

Equally notable are the results on training stability and robustness. The authors report that FTSformer substantially outperforms existing baselines not only in accuracy but also in the stability of its training process, an attribute they attribute to the Channel-Modulation Preprocessor’s balancing of channel signals before temporal modeling. They further evaluated robustness to exogenous perturbations, deliberately injecting noise or distortions into the external variables to simulate real-world conditions in which exogenous data sources may be incomplete, delayed, or corrupted. The model’s performance degraded more gracefully than that of competing methods, reinforcing the value of explicitly modeling the heterogeneity of external information rather than naively concatenating it into the input tensor. This robustness finding carries practical weight for financial institutions, where data pipelines are rarely pristine and a model that collapses when an indicator feed fails is of limited operational use.

The study situates itself within a rapidly evolving landscape of time series transformers. The original transformer architecture, introduced in 2017 for machine translation, was adapted to forecasting by models such as Informer and Autoformer, which introduced sparse attention and decomposition strategies for long sequences. More recently, iTransformer reoriented the attention mechanism across variables rather than time steps, while TimeXer explicitly empowered transformers with exogenous variables, a direct conceptual ancestor of FTSformer’s exogenous fusion philosophy. Meanwhile, finance-specific transformers such as Master have explored market-guided attention, and questions about whether transformers are even the right inductive bias for time series, posed pointedly by Zeng and colleagues in their DLinear work, continue to animate the field. FTSformer’s contribution is to combine the multi-scale decomposition trend with a principled, dual-domain treatment of exogenous heterogeneity, targeting the specific pathologies of financial data: high volatility, strong nonlinearity, and pronounced dependence on external drivers.

The authors are careful to frame the work as a flexible, exogenous-aware forecasting framework rather than a finance-only tool. The inclusion of the Canadian traffic dataset is a deliberate demonstration that the architecture’s benefits extend beyond markets, and the paper reports promising cross-domain transferability, suggesting that the dual-domain fusion and channel modulation principles could generalize to energy demand, weather, epidemiology, or any multivariate setting where external covariates drive the target. For practitioners in quantitative finance, the framework arrives amid growing interest in machine learning methods for asset pricing and risk management, following influential work such as Gu, Kelly, and Xiu’s empirical asset pricing study and the continuing proliferation of deep learning surveys in the field. The research was supported by the National Natural Science Foundation of China under Grant 62103177 and by the Key Technology Research and Development Program of Shandong Province under Grant 2025TSGCCZZB0893.

What makes the paper resonant beyond its benchmark numbers is its conceptual reframing of the forecasting problem. Rather than asking how best to extrapolate a single series, FTSformer treats prediction as an exercise in multi-source reasoning: the target’s own multi-scale history, the time-domain and frequency-domain signatures of external drivers, and the relative reliability of every input channel must all be weighed together, dynamically, at every forecasting step. As markets grow more interconnected and data feeds multiply, that framing is likely to influence how the next generation of forecasting systems is built. The study, received on 2 February 2026, accepted on 19 August 2026, and published on 30 August 2026, adds a technically detailed and empirically validated entry to that conversation, one that argues the future of accurate forecasting lies not in ever-larger single-series models but in architectures that know how to listen to the world around the signal.

At the heart of this research lies a dual empirical framework: a Difference-in-Differences (DID) model and a Double Machine Learning (DML) approach. These models are meticulously constructed to capture the nuanced economic impact of green finance policies, particularly their capacity to suppress corporate carbon emissions and concurrently improve key financial performance metrics over subsequent periods. The DID model evaluates the differential effects over time, while the DML model leverages advanced machine learning techniques to isolate causal relationships from complex datasets plagued by confounding variables.

The researchers focus on a crucial economic indicator denoted as (eco{n}_{it+1}), representing the economic benefits realized in the period following the implementation of green finance initiatives. This composite indicator encapsulates two vital dimensions: operating fees, which are the sum of administrative and selling expenses normalized by operating revenue, and corporate value, approximated through the ratio of book value to market value. These measures holistically reflect financial health and investor confidence post-adoption of environmentally conscious policies.

Underlying the modeling is the critical variable (\gamma ce_{it}), which mathematically represents the total inhibitory effect on direct and indirect carbon emissions attributable to green finance policy. Derived from earlier equations in the study, it translates policy measures into quantifiable emissions reductions. The coefficient (\varphi_0) captures how these emission reductions translate into economic gains, serving as the linchpin for measuring the policy’s effectiveness in fostering corporate sustainability and profitability.

The empirical findings are nothing short of compelling. Statistical evidence exhibits significant negative coefficients at the 1% significance level for (\gamma tce), (\gamma dce), and (\gamma ice), indicating that firms’ commitments to carbon emission reductions substantially lower operating fees while simultaneously enhancing corporate valuation. This dual impact suggests that environmentally driven operational efficiencies and improved market perceptions synergistically elevate firm performance.

Mechanistically, carbon reduction efforts translate into reduced waste management and material procurement costs. Firms achieve this through enhanced recycling, reuse initiatives, and leaner production processes, all of which directly lower operational expenses. This efficiency gain provides a compelling argument for corporate management to adopt green finance policies, not merely as a compliance requirement but as a core business strategy yielding measurable cost savings.

Beyond cost reductions, adopting eco-friendly practices positions firms favorably in the eyes of socially conscious investors and customers. The research cites prior literature emphasizing the growing importance of environmental, social, and governance (ESG) criteria in investment decisions. By demonstrating tangible progress in corporate carbon footprints, firms attract more capital inflows and consumer loyalty, both essential drivers of enhanced corporate value.

The integration of these models reveals the multifaceted benefits of green finance policy. Not only do these policies serve regulatory and normative functions by pushing firms toward carbon neutrality, but they also bolster economic incentives. This dual effect aligns environmental goals with investor and managerial interests, creating a virtuous cycle of sustainability and profitability.

Furthermore, the use of advanced machine learning methods in the DML framework introduces a new frontier in environmental economics. This approach improves the precision of estimating causal effects amidst complex and high-dimensional datasets, where traditional econometric methods might falter. Such technological incorporation heralds a new era for policy impact assessment in green finance.

From the policy perspective, the evidence supports intensified green finance initiatives as a lever for corporate transformation. Governments and financial institutions can amplify these benefits by designing instruments that directly incentivize emission reductions and reward operational efficiency. The study reinforces that careful calibration of these policies yields not only environmental but also significant economic dividends.

The research also bridges the gap between theory and practice by quantifying how emissions reductions translate into concrete financial outcomes. This empirical validation empowers firms with actionable data, making the business case for sustainability undeniably strong. By comprehensively capturing operating fees and corporate value metrics, the findings are robust and widely applicable across industries.

Moreover, the significant results at the 1% level underscore both the reliability and magnitude of the green finance policy effects. This statistical rigor lends credibility to the growing narrative that sustainability is not a burdensome mandate but a strategic opportunity. Firms’ carbon reduction pledges thus become a blueprint for financial success, fostered by policy incentives.

In addition, the research highlights the need for continuous refinement of green finance mechanisms. Understanding the distinct impacts on total, direct, and indirect carbon emissions allows policymakers to tailor interventions with surgical precision. This strategic insight can optimize resource allocation and enhance policy effectiveness.

The study’s multidimensional approach, combining classical difference analysis with cutting-edge machine learning, sets a methodological benchmark. It demonstrates how integrating diverse analytical tools can unravel complex socio-economic phenomena, paving the way for future investigations into green financial policies and corporate sustainability.

This research offers a timely and vital contribution to the discourse on sustainable finance, resonating with corporate leaders and policymakers alike. As the global economy charts a course toward net-zero emissions, understanding the economic benefits of green finance policies provides an essential compass for navigating this transformative shift.

Ultimately, the findings underscore that green finance is not merely a tool for environmental compliance but a potent catalyst for economic value creation. Firms that proactively engage with these policies stand to benefit through lowered costs, enhanced market valuations, and strengthened stakeholder trust, all critical to long-term competitiveness.

As corporations strive to balance profitability and responsibility, this study illuminates a path that reconciles these dual imperatives. The compelling empirical evidence offers a clear message: advancing corporate sustainability through green finance is both an environmental necessity and a profitable business strategy.

Subject of Research:

Article Title:

Article References:
Lyu, Y., Xiao, X. & Zhang, J. Green finance policy and corporate carbon emissions: advancing corporate sustainability. Humanit Soc Sci Commun 12, 1143 (2025). https://doi.org/10.1057/s41599-025-05197-w

Image Credits: AI Generated