KiT: A Foundation Model for Financial Time-Series Forecasting using DiffusionTransformers

Boyu Zhang; Haorui Li

cs.LGSep 28, 2026

KiT: A Foundation Model for Financial Time-Series Forecasting using DiffusionTransformers

Authors: Boyu Zhang, Haorui Li

Organizations: University of California, Los Angeles · Southeast University

Abstract

Financial candlestick forecasting is fundamental to quantitative investment, yet it remains exceptionally challenging due to extremely low signal-to-noise ratios and vast heterogeneity across markets and instruments. Existing approaches have largely attempted to introduce deep learning to capture hidden temporal features, but most adopt an auto-regressive formulation, which leads to error accumulation during inference. Meanwhile, general-purpose time-series foundation models are not tailored to the unique structure of k-line data and yield unsatisfactory performance on downstream candlestick forecasting tasks. To tackle these problems, we introduce KiT, a K-line Diffusion Transformer foundation model, and reformulate future prediction as conditional path generation via flow matching: given a historical context window, the model generates an ensemble of plausible future OHLCV trajectories. We pre-train KiT at multiple parameter scales on billions of candlestick bars spanning multiple markets and timescales. Across three markets and seven resolutions, KiT attains a mean return RankIC of 0.057 and a mean volatility RankIC of 0.66, leading at every timescale and outperforming both task-specific financial forecasters and general time-series foundation models. Code will be available at: https://github.com/Luciferbobo/KiT.

Figures & tables

Configuration	1m $\uparrow$	5m $\uparrow$	15m $\uparrow$	30m $\uparrow$	1h $\uparrow$	2h $\uparrow$	1d $\uparrow$	Mean $\uparrow$
KiT	$-$ .0215	$-$ .0576	$-$ .0526	$-$ .0474	$-$ .0383	$-$ .0651	$-$ .1168	$-$ .0571
Kronos-base	$-$ .0158	$-$ .0426	$-$ .0442	$-$ .0064	$-$ .0128	$-$ .0514	$-$ .1046	$-$ .0342
Kronos-base-FT	$-$ .0192	$-$ .0536	$-$ .0484	$-$ .0408	$-$ .0304	$-$ .0622	$-$ .1101	$-$ .0521
Sundial	$-$ .0148	$-$ .0386	$-$ .0434	$-$ .0082	$-$ .0046	$-$ .0214	$-$ .0837	$-$ .0209
Chronos-Bolt extended	$-$ .0362	$-$ .0088	$-$ .0186	$-$ .0964	$-$ .0068	$-$ .2159	$-$ .0052	$-$ .0461
Chronos-2	$-$ .0168	$-$ .0286	$-$ .0362	$-$ .0184	$-$ .0162	$-$ .0648	$-$ .0556	$-$ .0054

Table 1: Return forecasting by resolution. Evaluated across three markets in the test window. Each entry is the RankIC between the predicted mean terminal return and the realized close-to-close log-return sum.

Configuration	1m $\uparrow$	5m $\uparrow$	15m $\uparrow$	30m $\uparrow$	1h $\uparrow$	2h $\uparrow$	1d $\uparrow$	Mean $\uparrow$
KiT	$-$ .7599	$-$ .6947	$-$ .6900	$-$ .5919	$-$ .5958	$-$ .6652	$-$ .6225	$-$ .6600
Kronos-base	$-$ .5900	$-$ .4934	$-$ .3955	$-$ .3049	$-$ .4699	$-$ .5281	$-$ .4498	$-$ .4617
Kronos-base-FT	$-$ .6193	$-$ .5478	$-$ .4466	$-$ .3671	$-$ .4974	$-$ .5306	$-$ .4727	$-$ .4974
Sundial	$-$ .6024	$-$ .5477	$-$ .5164	$-$ .4591	$-$ .5691	$-$ .5534	$-$ .4486	$-$ .5281
Chronos-Bolt extended	$-$ .5766	$-$ .4821	$-$ .4113	$-$ .4289	$-$ .3543	$-$ .3797	$-$ .3310	$-$ .4234
Chronos-2	$-$ .4427	$-$ .3944	$-$ .3671	$-$ .4092	$-$ .4264	$-$ .4452	$-$ .2260	$-$ .3873

Table 2: Volatility prediction by resolution. Evaluated across three markets in the test window. Each entry is the RankIC between the predicted mean realized volatility and the realized sample volatility.

Appendix figures & tables9 assets

Supplementary material from the paper’s appendix.

Appendix

	KiT-S	KiT-M	KiT-L
Parameters	31.0M	101.0M	283.7M
Layers	8	13	21
Width	512	768	1,024
Heads	4	6	8
FFN	1,408	2,048	2,816
Steps	50k	50k	50k

Appendix

Table 3: Model configurations.

	KiT-S	KiT-M	KiT-L
Parameters	31.0M	101.0M	283.7M
Layers	8	13	21
Width	512	768	1,024
Heads	4	6	8
FFN	1,408	2,048	2,816
Steps	50k	50k	50k

Appendix

Table 3: Model configurations.

Resolution	History	Horizon
1 minute	1,200	120
5 minutes	960	96
15 minutes	800	64
30 minutes	640	40
1 hour	400	20
2 hours	360	20

Appendix

Table 4: Window size.

Configuration	1m $\uparrow$	5m $\uparrow$	15m $\uparrow$	30m $\uparrow$	1h $\uparrow$	2h $\uparrow$	1d $\uparrow$	Mean $\uparrow$
KiT	$-$ .0235	$-$ .0552	$-$ .0445	$-$ .0330	$-$ .0241	$-$ .0475	$-$ .1158	$-$ .0491
Kronos-base	$-$ .0072	$-$ .0344	$-$ .0368	$-$ .0244	$-$ .0209	$-$ .0186	$-$ .0048	$-$ .0067
Kronos-base-FT	$-$ .0164	$-$ .0506	$-$ .0392	$-$ .0136	$-$ .0125	$-$ .0348	$-$ .0058	$-$ .0247
Sundial	$-$ .0086	$-$ .0074	$-$ .0362	$-$ .0221	$-$ .0152	$-$ .0284	$-$ .1026	$-$ .0315
Chronos-Bolt extended	$-$ .0364	$-$ .0292	$-$ .0246	$-$ .0724	$-$ .0068	$-$ .2223	$-$ .0276	$-$ .0347
Chronos-2	$-$ .0188	$-$ .0383	$-$ .0372	$-$ .0402	$-$ .0841	$-$ .1789	$-$ .0962	$-$ .0161

Appendix

Table 5: Return IC by resolution. Higher is better; bold marks the best value in each column.

Configuration	1m $\uparrow$	5m $\uparrow$	15m $\uparrow$	30m $\uparrow$	1h $\uparrow$	2h $\uparrow$	1d $\uparrow$	Mean $\uparrow$
KiT	$-$ .7459	$-$ .6749	$-$ .6888	$-$ .6553	$-$ .6621	$-$ .7586	$-$ .5726	$-$ .6797
Kronos-base	$-$ .5654	$-$ .4369	$-$ .3616	$-$ .3175	$-$ .4491	$-$ .5745	$-$ .4377	$-$ .4490
Kronos-base-FT	$-$ .5836	$-$ .4695	$-$ .3884	$-$ .3759	$-$ .4667	$-$ .5736	$-$ .4470	$-$ .4721
Sundial	$-$ .6055	$-$ .5178	$-$ .5233	$-$ .5041	$-$ .5757	$-$ .5138	$-$ .4081	$-$ .5212
Chronos-Bolt extended	$-$ .5830	$-$ .4926	$-$ .4169	$-$ .4907	$-$ .4261	$-$ .4414	$-$ .1562	$-$ .4296
Chronos-2	$-$ .4156	$-$ .3688	$-$ .3185	$-$ .5068	$-$ .3665	$-$ .5365	$-$ .1272	$-$ .3771

Appendix

Table 6: Volatility IC by resolution. Higher is better; bold marks the best value in each column.

Configuration	IC/RankIC $\uparrow$
KiT	$\mathbf{0.0372/0.0407}$
Kronos-base	$0.0090/0.0148$
Kronos-base-FT	$0.0132/0.0205$

Appendix

Table 7: Price-series forecasting. Values are averaged over the seven resolutions and three markets.

	Identity $w_{\mathrm{id}}$ ( $w_{\mathrm{hist}}{=}1$ )			History $w_{\mathrm{hist}}$ ( $w_{\mathrm{id}}{=}1$ )
$w$	Prior $\downarrow$	Hist. $\downarrow$	Spread $\downarrow$	Prior $\downarrow$	Hist. $\downarrow$	Spread $\downarrow$
1	0.41	0.52	1.08	0.41	0.52	1.08
2	0.28	0.49	0.91	0.39	0.37	0.93
3	0.20	0.47	0.78	0.37	0.26	0.81
4	0.15	0.46	0.69	0.36	0.19	0.72

Appendix

Table 8: Guidance sweep averaged over 100 products.

Representation of a bar	Mean return RankIC $\uparrow$
Five-dimensional log-ratio state ( KiT )	0.0571
Close-only log-price input	0.0183
Naive OHLCV input (normalized)	0.0209

Appendix

Table 9: Ablation on the candlestick representation. Each row alters only the input state; all other settings follow the main model.

Configuration	Mean return RankIC $\uparrow$
AdaLN identity + per-bar calendar ( KiT )	0.0571
Identity $\rightarrow$ AdaLN
Market and sector only (drop instrument, scale)	0.0386
w/o identity	0.0329
Per-bar condition $\rightarrow$ token embeddings
w/o per-bar condition	0.0372

Appendix

Table 10: Ablation on conditioning injection. Identity denotes the window-constant signals (market, sector, instrument, timescale); Calendar denotes the per-bar signals (clock, session, day of week, month, day of year, event flags).

Explore similar work

EXAONE Forecast for Finance
Sep 7, 2026Seunghan Lee, Jaehoon Lee, Jun Seo +9Finance Benchmarks Convolutional
LG AI Research∗
Time Series as Language: A Universal Tokenizer for General-Purpose Time Series Foundation Models
May 31, 2026Yunhao Zhang, Ruiying Qi, Jiale Zheng +3Time Series Foundation Models Time Series
Shanghai Jiao Tong University · Huawei Noah’s Ark Lab
Crossing-Free Probabilistic K-Line Forecasts Without Retraining
Jul 29, 2026Runyao Yu, Yuchen Tao, Yujie Chen +2Probabilistic Forecasting Forecast
London Business School London, United Kingdom · RWTH Aachen University Aachen, Germany · The Chinese University of Hong Kong Shenzhen, China +1

Sep 7, 2026, cs.AIJ/K move · Enter open · S save

EXAONE Forecast for Finance

Seunghan Lee, Jaehoon Lee, Jun Seo, Tae Yoon Lim+8

LG AI Research∗

Finance Benchmarks Convolutional Text Corpora Inverse Design

Open paper AskPDF arXiv

This technical report presents EXAONE Forecast for Finance (EXAONE Finance), a financial time series (TS) foundation model (TSFM) tailored to financial forecasting. Recent TSFMs achieve strong zero-shot performance through large-scale pretraining. However, they are primarily developed for general-domain TS and largely rely on self-attention backbones whose computational cost grows quadratically with sequence length and variate count. Moreover, they assume fully observed inputs and are pretrained on corpora that fail to capture the unique dynamics of financial markets. These limitations hinder their applicability to finance, where long, many-channel, intermittently observed panels are common. To address these challenges, EXAONE Finance adopts an attention-free architecture, replacing self-attention with two simple yet effective linear-time operators: 1) a causal 1D convolution for temporal mixing and 2) a group-aware pooling multi-layer perceptron (MLP) for variate mixing. Furthermore, a masked context augmentation exposes the model to contiguous missing spans during training, improving robustness to the missingness pervasive in financial markets. EXAONE Finance is pretrained on a large-scale financial corpus covering not only equities but also foreign exchange, commodities, crypto-assets, fixed income, and macroeconomic indicators. On FinVerse, a financial forecasting benchmark covering diverse asset classes, EXAONE Finance attains state-of-the-art performance, ranking first across all three evaluation tiers---point-forecast accuracy, cross-sectional asset ranking, and portfolio profitability.

Abstract

Explore similar work

EXAONE Forecast for Finance

Time Series as Language: A Universal Tokenizer for General-Purpose Time Series Foundation Models

Crossing-Free Probabilistic K-Line Forecasts Without Retraining