Dynamic area coloring in plotly subplots - python-3.x

Question

I have 6 graphs done with plotly-subplots where I show the stock price and its simple moving average (20 days for instance). What I would like to do is for each subplot (except one), color the area in red where the current price is below the simple moving average and green if the price is above its simple moving average (and if possible as extra, no color when the stock price is crossing over or under the average: so the color area would start and end at the marker).
I understand I could use add_vrect to define my color area but this seems to be ok for static data: here each subplot would need its "own conditional coloring"
For clarity, I used only 3 stocks/graphs in my example (and the coloring would be on the second and third graphs):
import yfinance as yf
import pandas as pd
import numpy as np
from plotly.subplots import make_subplots
import plotly.graph_objects as go
tickers = ['AAPL', 'MSFT']
df = yf.download(tickers, period='1y', interval='1d', progress=False)['Close']
df_bench = yf.download('SPY', period='1y', interval='1d', progress=False)
df['AAPL_SMA'] = df['AAPL'].rolling(20).mean()
df['MSFT_SMA'] = df['MSFT'].rolling(20).mean()
fig = make_subplots(
rows=1, cols=3,
subplot_titles=('AAPL', 'MSFT')
)
fig.add_trace(
go.Candlestick(x=df_bench .index,
open=df_bench ['Open'],
high=df_bench ['High'],
low=df_bench ['Low'],
close=df_bench ['Close'],
name='Benchmark'),
row=1, col=1
)
fig.add_trace(
go.Scatter(x=df.index, y=df['AAPL'], name='AAPL', marker_color='blue'),
row=1, col=2
)
fig.add_trace(
go.Scatter(x=df.index, y=df['AAPL_SMA'], name='HYG/TLT_SMA', marker_color='red'),
row=1, col=2
)
fig.add_trace(
go.Scatter(x=df.index, y=df['MSFT'], name='MSFT', marker_color='blue'),
row=1, col=3
)
fig.add_trace(
go.Scatter(x=df.index, y=df['MSFT_SMA'], name='MSFT_SMA', marker_color='red'),
row=1, col=3
)
fig.update_layout(height=1200, width=2400, title_text='Dynamic coloring test',
showlegend=False, template='plotly_white',
hovermode='x unified', xaxis_rangeslider_visible=False)
fig.show()
I have added two booleans columns in my dataframe df where I test whether the area should be red (-1) or green (1):
df['AAPL_Sig'] = np.where(df['AAPL'] < df['AAPL_SMA'], -1, 1)
df['MSFT_Sig'] = np.where(df['MSFT'] < df['MSFT_SMA'], -1, 1)
Now I am stuck and would need some pointers as to how to use add_vrect (or maybe there are better functions -?-) dynamically and on some of the subplots.