Qlib HIST Configuration for CSI 300 Alpha360 Backtesting
Summary
This configuration specifies a Qlib workflow for training the HIST model on Alpha360 features for CSI 300 stocks and evaluating its signals in a portfolio backtest. The data handler normalizes features, fills missing feature values, drops rows without labels, and cross-sectionally rank-normalizes labels. Its label measures a close-to-close return over the next two trading sessions. The configuration separates training, validation, and test periods and uses a historical Chinese equity dataset.
For portfolio evaluation, it selects the top 50 ranked names and permits five holdings to be replaced at a time, with the CSI 300 index as benchmark. The backtest uses closing prices and specifies transaction costs and a price-limit threshold. Records include signal analysis and portfolio analysis. These settings describe an experiment, not its outcome: the document reports no performance, comparison, or robustness checks. Results would depend on the data, model implementation, and assumptions in the backtest, including execution at the close and the specified costs.
Key ideas
- The workflow trains Qlib's HIST model on Alpha360 features for CSI 300 constituents.
- Feature values are robustly normalized and missing feature values are filled.
- The label captures a forward close-to-close return across two sessions.
- The portfolio strategy selects 50 stocks and replaces up to five holdings at a time.
- The configuration specifies a benchmark, closing-price execution, transaction costs, and price limits.
- No empirical performance results or robustness analysis are provided.
Tags
Full text
# workflow_config_hist_Alpha360.yaml
```yaml
qlib_init:
provider_uri: "~/.qlib/qlib_data/cn_data"
region: cn
market: &market csi300
benchmark: &benchmark SH000300
data_handler_config: &data_handler_config
start_time: 2008-01-01
end_time: 2020-08-01
fit_start_time: 2008-01-01
fit_end_time: 2014-12-31
instruments: *market
infer_processors:
- class: RobustZScoreNorm
kwargs:
fields_group: feature
clip_outlier: true
- class: Fillna
kwargs:
fields_group: feature
learn_processors:
- class: DropnaLabel
- class: CSRankNorm
kwargs:
fields_group: label
label: ["Ref($close, -2) / Ref($close, -1) - 1"]
port_analysis_config: &port_analysis_config
strategy:
class: TopkDropoutStrategy
module_path: qlib.contrib.strategy
kwargs:
signal: <PRED>
topk: 50
n_drop: 5
backtest:
start_time: 2017-01-01
end_time: 2020-08-01
account: 100000000
benchmark: *benchmark
exchange_kwargs:
limit_threshold: 0.095
deal_price: close
open_cost: 0.0005
close_cost: 0.0015
min_cost: 5
task:
model:
class: HIST
module_path: qlib.contrib.model.pytorch_hist
kwargs:
d_feat: 6
hidden_size: 64
num_layers: 2
dropout: 0
n_epochs: 200
lr: 1e-4
early_stop: 20
metric: ic
loss: mse
base_model: LSTM
model_path: "benchmarks/LSTM/model_lstm_csi300.pkl"
stock2concept: "benchmarks/HIST/qlib_csi300_stock2concept.npy"
stock_index: "benchmarks/HIST/qlib_csi300_stock_index.json"
GPU: 0
dataset:
class: DatasetH
module_path: qlib.data.dataset
kwargs:
handler:
class: Alpha360
module_path: qlib.contrib.data.handler
kwargs: *data_handler_config
segments:
train: [2008-01-01, 2014-12-31]
valid: [2015-01-01, 2016-12-31]
test: [2017-01-01, 2020-08-01]
record:
- class: SignalRecord
module_path: qlib.workflow.record_temp
kwargs:
model: <MODEL>
dataset: <DATASET>
- class: SigAnaRecord
module_path: qlib.workflow.record_temp
kwargs:
ana_long_short: False
ann_scaler: 252
- class: PortAnaRecord
module_path: qlib.workflow.record_temp
kwargs:
config: *port_analysis_config
```Shown in full with attribution under the source's licence. Licence: MIT
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.