Fear and Greed indicators

AIM: Find indicators to estimate the current market sentiment, and for each one curate a Perplexity Page commenting the rationale and critics about it.

import os
import dotenv # pip install python-dotenv
import pandas as pd

import yfinance as yf # pip install yfinance
from fredapi import Fred # pip install fredapi

dotenv.load_dotenv()

print("yfinance version:", yf.__version__)

yfinance version: 0.2.65

F&G Indicator #1: margin borrowing

https://www.perplexity.ai/page/margin-debt-as-percentage-of-m-9lQvy1AIRXC1pvm0J45Z3A

sp500tr_raw = yf.Ticker('^SP500TR').history(period='max', auto_adjust=True)
display(sp500tr_raw)
sp500tr_ss = sp500tr_raw["Close"]
sp500tr_ss.index = sp500tr_ss.index.date

	Open	High	Low	Close	Volume	Dividends	Stock Splits
Date
1988-01-04 00:00:00-05:00	256.019989	256.019989	256.019989	256.019989	0	0.0	0.0
1988-01-05 00:00:00-05:00	258.769989	258.769989	258.769989	258.769989	0	0.0	0.0
1988-01-06 00:00:00-05:00	259.029999	259.029999	259.029999	259.029999	0	0.0	0.0
1988-01-07 00:00:00-05:00	261.209991	261.209991	261.209991	261.209991	0	0.0	0.0
1988-01-08 00:00:00-05:00	243.550003	243.550003	243.550003	243.550003	0	0.0	0.0
...	...	...	...	...	...	...	...
2025-10-20 00:00:00-04:00	14838.320312	14958.740234	14838.320312	14938.290039	0	0.0	0.0
2025-10-21 00:00:00-04:00	14941.929688	14975.950195	14910.530273	14938.830078	0	0.0	0.0
2025-10-22 00:00:00-04:00	14952.349609	14953.259766	14762.370117	14859.330078	0	0.0	0.0
2025-10-23 00:00:00-04:00	14868.969727	14970.639648	14861.750000	14946.129883	0	0.0	0.0
2025-10-24 00:00:00-04:00	15021.400391	15098.990234	15021.400391	15064.910156	0	0.0	0.0

9526 rows × 7 columns

fred = Fred(api_key=os.getenv("FRED_API_KEY"))
m2_raw = fred.get_series('M2SL')
display(m2_raw)
m2_ss = pd.Series(m2_raw)
m2_ss.index = pd.to_datetime(m2_ss.index)

1959-01-01      286.6
1959-02-01      287.7
1959-03-01      289.2
1959-04-01      290.1
1959-05-01      292.2
               ...   
2025-04-01    21805.1
2025-05-01    21884.6
2025-06-01    22022.1
2025-07-01    22115.8
2025-08-01    22195.4
Length: 800, dtype: float64

margin_raw = pd.read_excel('https://www.finra.org/sites/default/files/2021-03/margin-statistics.xlsx')
display(margin_raw)
margin_ss = margin_raw.set_index('Year-Month')["Debit Balances in Customers' Securities Margin Accounts"]
margin_ss.index = pd.to_datetime(margin_ss.index)

	Year-Month	Debit Balances in Customers' Securities Margin Accounts	Free Credit Balances in Customers' Cash Accounts	Free Credit Balances in Customers' Securities Margin Accounts
0	2025-09	1126494	204106	194884.0
1	2025-08	1059723	188221	181563.0
2	2025-07	1022548	194712	186672.0
3	2025-06	1007961	200573	185584.0
4	2025-05	920960	187685	176698.0
...	...	...	...	...
340	1997-05	111097	66879	NaN
341	1997-04	102842	64825	NaN
342	1997-03	104835	69553	NaN
343	1997-02	103886	67058	NaN
344	1997-01	103337	68856	NaN

345 rows × 4 columns

df = (
    sp500tr_ss
    .div(sp500tr_ss.iloc[-1]/10) # rescale to be 10 as last value 
    .rename("S&P 500 Total Returns")
    .to_frame()
    .join(margin_ss.div(1e6).rename("Margin (T$)")) # raw is in M$
    .join(m2_ss.div(1e3).rename("M2 (T$)")) # raw is in B$
    .ffill()
    .assign(**{
        "%Margin/M2": lambda x: x["Margin (T$)"] / x["M2 (T$)"] * 100
    })
)
df

	S&P 500 Total Returns	Margin (T$)	M2 (T$)	%Margin/M2
1988-01-04	0.169945	NaN	NaN	NaN
1988-01-05	0.171770	NaN	NaN	NaN
1988-01-06	0.171943	NaN	NaN	NaN
1988-01-07	0.173390	NaN	NaN	NaN
1988-01-08	0.161667	NaN	NaN	NaN
...	...	...	...	...
2025-10-20	9.915950	1.059723	22.1954	4.774516
2025-10-21	9.916309	1.059723	22.1954	4.774516
2025-10-22	9.863537	1.059723	22.1954	4.774516
2025-10-23	9.921154	1.059723	22.1954	4.774516
2025-10-24	10.000000	1.059723	22.1954	4.774516

9526 rows × 4 columns

TITLE = "Fear&Greed indicator #1: % Margin debt over M2"
INCLUDE = ["S&P 500 Total Returns", "%Margin/M2"]
# plotly
import plotly.express as px # pip install plotly
fig = px.line(
    df.dropna()[INCLUDE].reset_index().melt(id_vars="index"),
    x="index",
    y="value",
    color="variable",
    title=TITLE,
).update_layout(
    yaxis_type="log",
    width=800,
).show()
# matplotlib
import matplotlib.pyplot as plt
df_plot = df.dropna()[INCLUDE]
fig, ax = plt.subplots(figsize=(10, 5))
for col in df_plot.columns:
    ax.plot(df_plot.index, df_plot[col], label=col)
ax.set_yscale('log')
ax.set_title(TITLE)
ax.legend()
ax.grid()
plt.show()