For flexible, fast, low cost queries
Data done for you
This is a method whereby you provide the api with any price data you wish for rapid turnaround and much lower credits used.
Cointegration (EG) helps to identify if there is a long term mean reverting stationary relationship for a given pair of assets. EG stands for Engle Granger. This is the test used to check for cointegration.
To determine cointegration, we expect that the t-stat < the critical value (cv) in addition to the p-value being less than 0.05. If this criteria is met, then is_coint will be flagged as true.
If inc_trend shows as true, this simply means that models fit the derived spread better when a trend is assumed. If looking for a stationary timeseries (oscillations up and down around a mean with mean, variance and autocovariance constant irrespective of time), then a trend may not be optimal.
All price series must match in length
curl -X POST "https://api.cryptowizards.net/v1beta/cointegration" \
-H "Content-Type: application/json" \
-H "X-api-key: REPLACE_WITH_YOUR_API_KEY" \
-d '{
"series_1_closes": [1.1, 1.2, 1.3, 1.4],
"series_2_closes": [4.1, 4.2, 4.3, 4.4],
"spread_type": "Static",
"roll_w": 42,
"with_history": false
}'
{
"data": {
"p_value": 0.9095462189033201,
"t_stat": -1.2046475596299255,
"cv": -3.498585183773327,
"is_coint": false,
"inc_trend": true
},
"history": null
}
{
"data": {
"p_value": 0.9095462189033201,
"t_stat": -1.2046475596299255,
"cv": -3.498585183773327,
"is_coint": false,
"inc_trend": true
},
"history": {
"spread": [
-0.048293964169238146,
-0.04606073255866372,
...
],
"zscore": [
-0.36609003525530315,
-0.349161132169538,
...
],
"zscore_roll": [
0.0,
...
-2.1615087099114954
],
"hedge_ratio": 0.25128116258971733,
"half_life": 6.442774277370148,
"hurst": 1.0051152158260312,
"sigma0crossings": 7,
"sigma2crossings": 2,
"log_used": true,
"last_zscore": {
"zscore": -2.3438770448984756, // (normalised spread)
"zscore_roll": -2.1615087099114954
}
}
}