Order book from Binance in R

Last update: 2023-06-10

GET depth function

The binance_depth(api = “spot”, pair, quiet = TRUE) function takes as three arguments:

api character (mandatory, default: spot): specify the api module to use. It can be:
- spot: for Spot market;
- fapi: for Futures and Margin markets denominated in USD, denoted in the website by USD-M;
- dapi: for Futures and Margin markets denominated in the underlying asset, denoted in the website by COIN-M;
- eapi: for Options market.
pair character (optional, default: BTCUSDT): the a trading pair available on Binance.com, for example BTCUSDT, ETHUSDT, ETHBTC.
quiet logical (mandatory, default: FALSE): if TRUE the function will not display messages and warnings.

Example: Spot (api)

# import depth data from binance 
depth <- binance_depth(api = "spot", pair = "BTCUSDT", quiet = TRUE)

data <- dplyr::bind_rows(
  tail(dplyr::filter(depth, side == "ASK"), n = 3), # best 3 ask
  head(dplyr::filter(depth, side == "BID"), n = 3)  # best 3 bid
) 

format_kable(data, caption)

Last Update: 10 Jun 2023 at 03:23 CET
date	market	pair	side	price	quantity
2023-06-10 03:23:21	spot	BTCUSDT	ASK	26459.09	0.02288
2023-06-10 03:23:21	spot	BTCUSDT	ASK	26459.08	0.02138
2023-06-10 03:23:21	spot	BTCUSDT	ASK	26459.07	10.05993
2023-06-10 03:23:21	spot	BTCUSDT	BID	26459.06	1.66527
2023-06-10 03:23:21	spot	BTCUSDT	BID	26459.05	0.00946
2023-06-10 03:23:21	spot	BTCUSDT	BID	26459.01	0.03392

Example: Futures USD-M (fapi)

# import depth data from binance 
depth_fapi <- binance_depth(api = "fapi", pair = "BTCUSDT", quiet = TRUE)

dplyr::bind_rows(
  tail(dplyr::filter(depth_fapi, side == "ASK"), n = 3), # best 3 ask
  head(dplyr::filter(depth_fapi, side == "BID"), n = 3)  # best 3 bid
) %>% 
  format_kable(caption)

Last Update: 10 Jun 2023 at 03:23 CET
date	market	pair	side	price	quantity
2023-06-10 03:23:22	usd-m	BTCUSDT	ASK	26444.1	0.001
2023-06-10 03:23:22	usd-m	BTCUSDT	ASK	26444.0	0.016
2023-06-10 03:23:22	usd-m	BTCUSDT	ASK	26443.9	17.520
2023-06-10 03:23:22	usd-m	BTCUSDT	BID	26443.8	14.344
2023-06-10 03:23:22	usd-m	BTCUSDT	BID	26443.7	0.034
2023-06-10 03:23:22	usd-m	BTCUSDT	BID	26443.6	0.489

Example: Futures COIN-M (dapi)

# import depth data from binance 
depth_dapi <- binance_depth(api = "dapi", pair = "BTCUSD_PERP", quiet = TRUE)

dplyr::bind_rows(
  tail(dplyr::filter(depth_dapi, side == "ASK"), n = 3), # best 3 ask
  head(dplyr::filter(depth_dapi, side == "BID"), n = 3)  # best 3 bid
) %>% 
  format_kable(caption)

Last Update: 10 Jun 2023 at 03:23 CET
date	market	pair	side	price	quantity
2023-06-10 03:23:24	coin-m	BTCUSD_PERP	ASK	26451.0	926
2023-06-10 03:23:24	coin-m	BTCUSD_PERP	ASK	26450.9	3
2023-06-10 03:23:24	coin-m	BTCUSD_PERP	ASK	26450.7	17402
2023-06-10 03:23:24	coin-m	BTCUSD_PERP	BID	26450.6	425
2023-06-10 03:23:24	coin-m	BTCUSD_PERP	BID	26450.5	538
2023-06-10 03:23:24	coin-m	BTCUSD_PERP	BID	26450.4	55

Example: Options (eapi)

df_options <- binance_exchange_info(api = "eapi", pair = "BTCUSDT")

symbol <- filter(df_options, strikePrice == "27000.00000000")[1,]$symbol

df_options <- binance_exchange_info(api = "eapi", pair = "BTCUSDT")
symbol <- filter(df_options, as.numeric(strikePrice) == trunc(data$price[1]/1000)*1000)[1,]$symbol

# import depth data from binance 
depth_eapi <- binance_depth(api = "eapi", pair = symbol, quiet = TRUE)

dplyr::bind_rows(
  tail(dplyr::filter(depth_eapi, side == "ASK"), n = 3), # best 3 ask
  head(dplyr::filter(depth_eapi, side == "BID"), n = 3)  # best 3 bid
) %>% 
  format_kable(caption)

Last Update: 10 Jun 2023 at 03:23 CET
date	market	pair	side	price	quantity
2023-06-10 03:23:25	options	BTC-230610-26000-C	ASK	800	0.02
2023-06-10 03:23:25	options	BTC-230610-26000-C	ASK	660	0.05
2023-06-10 03:23:25	options	BTC-230610-26000-C	ASK	590	2.50
2023-06-10 03:23:25	options	BTC-230610-26000-C	BID	325	2.50

Example: Micro price and Imbalance

Denoting by:

$P^{\text{BID}}_t$ : best BID price.
$P^{\text{ASK}}_t$ : best ASK price.
$Q^{\text{BID}}_t$ : quantity as limit buy order at BID price.
$Q^{\text{ASK}}_t$ : quantity as limit sell order at ASK price.

The mid-price is computed as simple mean among the best BID-ASK price:

$P^{\text{Mid}}_t = \frac{P^{\text{BID}}_t + P^{\text{ASK}}_t}{2}$

The bid-ask spread is the mean distance among the best BID and ASK price:

$S_t = \frac{P^{\text{ASK}}_t - P^{\text{BID}}_t}{2}$

The microprice is defined as weighted sum considering also the quantities $H$ at the best BID-ASK:

$P^{\text{Micro}}_t = \frac{P^{\text{BID}}_t \cdot Q^{\text{BID}}_t + P^{\text{ASK}}_t \cdot Q^{\text{ASK}}_t}{Q^{\text{BID}}_t + Q^{\text{ASK}}_t}$ The order imbalance is defined as the proportion of quantity on the best BID with respect to all the quantities:

$I_t = \frac{Q^{\text{BID}}_t}{Q^{\text{BID}}_t + Q^{\text{ASK}}_t}$ - Gatheral and Oomen (2009)

# best ask and bid 
bask <- tail(dplyr::filter(depth, side == "ASK"), n = 1)
bbid <- head(dplyr::filter(depth, side == "BID"), n = 1)

# compute mid-price 
mid_price  <- (bbid$price + bask$price)/2

# compute bid-ask spread 
bid_ask_spread  <- (bask$price - bbid$price)/2

# compute micro-price 
micro_price  <- (bbid$price*bbid$quantity + bask$price*bask$quantity)/(bbid$quantity + bask$quantity)

# compute order-imbalance 
bid_imbalance <- (bbid$quantity)/(bask$quantity + bbid$quantity)*100

df_mp_example <- dplyr::tibble(
        best_ask = bask$price,
        ask = bask$quantity, 
        mid_price = mid_price,
        micro_price = micro_price,
        bid_ask_spread = bid_ask_spread,
        imbalance = paste0(round(bid_imbalance,2), "%"),
        bid = bbid$quantity,
        best_bid = bbid$price) 

colnames(df_mp_example) <- c("$P^{\\text{ASK}}_t$", "$\\text{ASK}$", "$P^{\\text{Mid}}_t$", "$P^{\\text{Micro}}_t$",
                             "$S_t$", "$I_t$", "$\\text{BID}$", "$P^{\\text{BID}}_t$")
                        
df_mp_example %>%
  format_kable(caption)

Last Update: 10 Jun 2023 at 03:23 CET
$P^{\text{ASK}}_t$	$\text{ASK}$	$P^{\text{Mid}}_t$	$P^{\text{Micro}}_t$	$S_t$	$I_t$	$\text{BID}$	$P^{\text{BID}}_t$
26459.07	10.05993	26459.06	26459.07	0.005	14.2%	1.66527	26459.06

Build an order-book from Depth

An order book is a table with three columns: - price: the price level - bid: the quantity that is on the buy side - ask: the quantity that is on the sell side

We have also added three more columns:

buy: the quantity bought at that level
sell: the quantity sold at that level
net: the balance between buy and sell at that level, computed as $\text{net} = \text{buy} - \text{sell}$

OrderBook

The function OrderBook has the purpose of structuring a dataset in the same structure as this Table.

OrderBook(data = depth, from = micro_price*0.99, to = micro_price*1.01, levels = 10, as_datatable = TRUE)

	price	bid	ask
1	26,723.66	0.00	179.08
2	26,664.86	0.00	60.53
3	26,606.06	0.00	49.97
4	26,547.27	0.00	54.27
5	26,488.47	74.53	85.18
6	26,429.67	77.16	0.00
7	26,370.87	40.76	0.00
8	26,312.07	69.53	0.00
9	26,253.28	109.45	0.00
10	26,194.48	38.30	0.00

Showing 1 to 10 of 10 entries

Example: Liquidity curve

# order by price (descending)
plot_data <- dplyr::arrange(depth, price)
# compute cumulated quantity 
plot_data$cumQ <- cumsum(plot_data$quantity)

mp = c(x = micro_price, xend = micro_price, 
       y = 0, yend = plot_data$cumQ[which(plot_data$side == "ASK")[1]])

ggplot()+
  geom_line(data = plot_data, aes(price, cumQ, color = side))+
  geom_segment(aes(x = mp[1], xend = mp[2], y = mp[3], yend = mp[4]), linetype="dashed", color = "red")+
  geom_point(aes(x = mp[1], y = mp[4]), size=2, color = "black")+
  geom_label(aes(x = mp[1], y = mp[4] + mp[4]/10, label = round(micro_price, 2)), size=2, color = "black")+
  scale_color_manual(values = c(ASK = "red", BID = "green"))+
  labs(x = "Price", y = "Cumulated Quantity", color = "Side", caption = caption)+
  theme(legend.position = "top")+
  theme(axis.title  = element_text(face = "bold"), 
        plot.title  = element_text(face = "bold"),
        axis.line = element_line(),
        axis.text.x = element_text(angle = 0, face = "bold", size = 7), 
        axis.text.y = element_text(face = "bold"), 
        axis.title.x = element_text(face = "bold",size = 10),
        axis.title.y = element_text(face = "bold", size = 10),
        plot.subtitle = element_text(face = "italic"),
        plot.caption = element_text(face = "italic"),
        panel.grid.minor.x = element_blank(),
        panel.grid.minor.y = element_blank(),
        panel.grid.major.x = element_blank(),
        panel.grid.major.y = element_blank(),
        panel.border = element_blank(),
        strip.background = element_blank(),
        panel.background = element_blank(),
        strip.text = element_text(angle = 0, face = "bold", size = 7), 
        legend.title = element_text(face = "bold", size = 11),
        legend.text = element_text(face = "italic", size = 10),
        legend.box.background = element_rect(),
        legend.position = "top") 

Example: Depth curve

# ASK orders 
ask <- dplyr::filter(depth, side == "ASK") 
ask <- dplyr::arrange(ask, price) # arrange by price (descending) 
ask$cumQ <- cumsum(ask$quantity) # cumulated quantity 

# BID orders 
bid <- dplyr::filter(depth, side == "BID")
bid <- dplyr::arrange(bid, dplyr::desc(price)) # arrange by price (ascending) 
bid$cumQ <- cumsum(bid$quantity) # cumulated quantity 

plot_data <- dplyr::bind_rows(ask, bid)

mp = c(x = micro_price, xend = micro_price, 
       y = 0, yend = 0)

ggplot()+
  geom_line(data = plot_data, aes(price, cumQ, color = side))+
  geom_label(aes(x = mp[1], y = mp[4], label = round(micro_price, 2)), size=2, color = "black")+
  scale_color_manual(values = c(ASK = "red", BID = "green"))+
  labs(x = "Price", y = "Quantity", color = "Side", caption = caption)+
  theme(legend.position = "top")+
  theme(axis.title  = element_text(face = "bold"), 
        plot.title  = element_text(face = "bold"),
        axis.line = element_line(),
        axis.text.x = element_text(angle = 0, face = "bold", size = 7), 
        axis.text.y = element_text(face = "bold"), 
        axis.title.x = element_text(face = "bold",size = 10),
        axis.title.y = element_text(face = "bold", size = 10),
        plot.subtitle = element_text(face = "italic"),
        plot.caption = element_text(face = "italic"),
        panel.grid.minor.x = element_blank(),
        panel.grid.minor.y = element_blank(),
        panel.grid.major.x = element_blank(),
        panel.grid.major.y = element_blank(),
        panel.border = element_blank(),
        strip.background = element_blank(),
        panel.background = element_blank(),
        strip.text = element_text(angle = 0, face = "bold", size = 7), 
        legend.title = element_text(face = "bold", size = 11),
        legend.text = element_text(face = "italic", size = 10),
        legend.box.background = element_rect(),
        legend.position = "top") 

Example: Quantile detection of supports and resitances

ask <- filter(depth, side == "ASK")
ob <- OrderBook(ask) 
ob$ask <- ob$ask/sum(ob$ask)
ob$col <- ifelse(ob$ask > quantile(ob$ask, 0.8), "up", "dw")

ggplot(ob)+
  geom_bar(data = ob, stat = "identity", aes(price, ask, fill = col), color = "black") +
  geom_label(data = filter(ob, col == "up"), aes(price, ask, label = round(price)), color = "black") +
  coord_flip()+
  scale_x_continuous(breaks = round(ob$price))+
  scale_fill_manual(values = c(up = "red", dw = "gray"))+
  labs(x = "Price", y = "ASK Quantity", caption = caption)+
  theme(axis.title  = element_text(face = "bold"), 
        plot.title  = element_text(face = "bold"),
        axis.line = element_line(),
        axis.text.x = element_text(angle = 0, face = "bold", size = 7), 
        axis.text.y = element_text(face = "bold"), 
        axis.title.x = element_text(face = "bold",size = 10),
        axis.title.y = element_text(face = "bold", size = 10),
        plot.subtitle = element_text(face = "italic"),
        plot.caption = element_text(face = "italic"),
        panel.grid.minor.x = element_blank(),
        panel.grid.minor.y = element_blank(),
        panel.grid.major.x = element_blank(),
        panel.grid.major.y = element_blank(),
        panel.border = element_blank(),
        strip.background = element_blank(),
        panel.background = element_blank(),
        strip.text = element_text(angle = 0, face = "bold", size = 7), 
        legend.title = element_text(face = "bold", size = 11),
        legend.text = element_text(face = "italic", size = 10),
        legend.box.background = element_rect(),
        legend.position = "none") 

bid <- filter(depth, side == "BID")
ob <- OrderBook(bid) 
ob$bid <- ob$bid/sum(ob$bid)
ob$col <- ifelse(ob$bid > quantile(ob$bid, 0.8), "up", "dw")

ggplot(ob)+
  geom_bar(data = ob, stat = "identity", aes(price, bid, fill = col), color = "black") +
  geom_label(data = filter(ob, col == "up"), aes(price, bid, label = round(price)), color = "black") +
  labs(x = "Price", y = "BID Quantity", caption = caption)+
  coord_flip()+
  scale_x_continuous(breaks = round(ob$price))+
  scale_fill_manual(values = c(up = "green", dw = "gray"))+
  theme(axis.title  = element_text(face = "bold"), 
        plot.title  = element_text(face = "bold"),
        axis.line = element_line(),
        axis.text.x = element_text(angle = 0, face = "bold", size = 7), 
        axis.text.y = element_text(face = "bold"), 
        axis.title.x = element_text(face = "bold",size = 10),
        axis.title.y = element_text(face = "bold", size = 10),
        plot.subtitle = element_text(face = "italic"),
        plot.caption = element_text(face = "italic"),
        panel.grid.minor.x = element_blank(),
        panel.grid.minor.y = element_blank(),
        panel.grid.major.x = element_blank(),
        panel.grid.major.y = element_blank(),
        panel.border = element_blank(),
        strip.background = element_blank(),
        panel.background = element_blank(),
        strip.text = element_text(angle = 0, face = "bold", size = 7), 
        legend.title = element_text(face = "bold", size = 11),
        legend.text = element_text(face = "italic", size = 10),
        legend.box.background = element_rect(),
        legend.position = "none") 