What is the probability that the inflation rate

worldwide will increase in the next year based

on the Bayes Theorem?

Math Analysis and Interpretations HL

 Table of contents:

Introduction................................................................................4
Definition of Bayes theorem....................................................5
Economic indicators of inflation:.............................................6
Mathematical model.................................................................. 9
Problem formulation:............................................................... 9
Empirical study:....................................................................... 11
Calculation............................................................................. 12
Evaluation.................................................................................14
Limitations............................................................................. 14
Improvements........................................................................ 14
Conclusion................................................................................ 15
Sources:..................................................................................16
Introduction

How much can inflation increase in two years? How rapidly has inflation increased in the
past? In 2022, in France, the price of milk increased by 17-20% in only a year. So, why do prices
increase over time? This concept is known as inflation. Inflation is a broad economic term,
describing rising prices of goods and services over time, leading to a decrease in purchasing
power of money. Inflation can happen in different ways and at different speeds. For example, the
Great Depression was a great example of what happens when inflation increases significantly in
a short period. This phenomenon is called hyperinflation, though hyperinflation is rare. The
opposite of this phenomenon is called deflation, when prices of goods and services get lower
over time. To measure inflation. Though a moderate level of inflation is normal in a healthy
economy, if inflation increases too quickly or becomes unpredictable, it can create problems for
both consumers and businesses. Two ways to measure inflation are to use the month-over-month
consumer price index (CPI), which measures the change in the price of goods and services that a
consumer pays for, or to use the producer price index (PPI), which measures the change of price
in production costs.

In this project, I aim to calculate the probability of an increase in inflation in the next year
using the Bayes theorem. Bayles theorem is a mathematical formula commonly used to calculate
conditional probabilities. Conditional probability is the chance of an event occurring based on a
previously stated event. Bayes’ theorem yields a way to update probabilities given new or
additional information. This theorem can be applied in different fields, including finance, where
it can be used to calculate the rate of risk of lending money to a borrower. In my case, Bayes's
theorem will be fundamental to prove my hypothesis by taking into account the criteria that
affect the increase of inflation in the next year.

The reason I chose this topic is because I plan on majoring in business management and/
or finance in the future, plus, I am highly familiar with the topic of inflation due to us having a
whole dedicated unit about inflation last year in DP1. In the past, I have worked as an employee
for my mother’s business and have found enjoyment in learning about the economy and the
concept of managing money. Additionally, the lesson that we were taught last year in MATH AI
DP1 about probability and the Bayes theorem I found was interesting and easy to understand,
and found it a powerful tool to determine uncertainty in a real-life situation. For example,
guessing the winner of the next football game.

In the following sections, the second section will be dedicated to defining the conditional
probabilities, including Bayes theorem. The third section will explain factors influencing
inflation worldwide, and economic indicators linked to inflation. The fourth section will be
devoted to providing mathematical models based on Bayes theorem. Finally, to assess our model
chosen, I'll use collected data from the “Federal Reserve Economic Data.”
Definition of Bayes theorem.
Bayes theorem, introduced by the mathematician Thomas Bayes provides a method to
determine probabilities or determine parameters of probability distributions based on
observations. In other words, Bayes gives us a method to calculate conditional probabilities. The
Bayes theorem starts with a prior probability that certain elements of the world hold, then yields
observations needed to modify that probability to work out the posterior probability.
Mathematically, Bayes theorem is defined as follows: let 𝐵1 ,𝐵2 ,𝐵3 … 𝐵𝑛 be n possible events

happening in the world. The prior probability that 𝐵𝑖 occurs is denoted as P(𝐵𝑖 ). Let A be any

event having a probability P(A|𝐵𝑖 ) to be realized given that 𝐵𝑖 occurs. Then, according to the
total law probability, the overall probability that the event “A” occurs is:

P(A) = P(𝐴 ∩ 𝐵1 ) + P(𝐴 ∩ 𝐵2 ) + … + P(𝐴 ∩ 𝐵𝑛)

According to the conditional probability formula:

𝑃(𝐴∩𝐵𝑖 )
P(A|𝐵𝑖 ) =
𝑃(𝐵𝑖 )

which means 𝑃(𝐴 ∩ 𝐵𝑖 ) = P(A|𝐵𝑖 ) · P(𝐵𝑖 )

Therefore: 𝑃(𝐴) = P(A|𝐵1 ) · P(𝐵1 ) + P(A|𝐵2 ) · P(𝐵2 ) + … + P(A|𝐵𝑛) · P(𝐵𝑛)

The conditional probability of 𝐵𝑖 given that “A” has occurred is the posterior probability which is
defined by:

𝑃(𝐴∩𝐵𝑖 )
P(𝐵𝑖 |𝐴) = 𝑃(𝐴)

Bayes theorem is an important tool in statistical analysis across various economic

domains, including finance, it also improves the decision-making process allowing us to predict
events more accurately. In a real-world scenario, the Bayes theorem is often used by businesses
to predict future purchases based on the behavior of past customers, which benefits the business
by finding the right products to recommend to past and future customers while improving the
shopping experience for all. In a more Mathematical, financial sense, Bayes theorem is used by
analysts to predict future market trends and stock price movement based on the impact of current
market conditions and historical data facilitating decisions. alternatively, the Bayesian formula
can also be used in mathematical, and economic research to update for example the probability
of a hypothesis becoming true as more and more information is given. This in a general sense,
helps refine models based on new outcomes. (GeeksforGeeks)

Economic indicators of inflation:

To study inflation and price stability the researchers as well as stakeholders have used
some economic indicators that have shown a correlative relationship with inflation such as CPI,
globalization and trade, the growth of GDP and imports… These indicators allow us to create a
quantitative approach for the prediction of worldwide inflation. This indicator should be taken
into account in the prediction process.

More specifically Carvalho et al (2018) investigated that the persistence of inflation,

globalization, and trade and the level of openness to trade is positively correlated with inflation.
Lim and Sek (2015) proved that the growth of GDP and imports remarkably impact inflation in
low-inflation nations. On the other hand, money supply, national spending, and GDP influence
inflation in high-inflation nations. On the other hand, Adjemian et al. (2023), who tried to
understand the sudden rise in food prices found that agricultural commodity prices and exchange
rates have significant roles in defining both short and long-term fluctuations in US food prices.
Eftekhari and Kiaee (2015) demonstrated that money growth, GDP, oil price, and the level of
income, are strong indicators regarding the inflation rate. (Yilmaz, Sevgi Coskun)
Figure 1. Correlation between economic indicators and CPI

Jai Kasera, Howard Powers (2022) used the Month over Month Consumer Price Index
Seasonally Adjusted (CPI) and found variables who have a strong correlation with economic
indicators and CPI. This correlation is displayed in Figure 1.
Economic indicators. definitions Acronym
used

CPI The average change over time in the prices paid

by consumers 𝐸1

PPI indexes that measures the average change over

time in selling prices received by domestic 𝐸2
producers of goods and services

GDP monetary value of all finished goods and services

made within a country during a specific period. 𝐸3

International Imports buying and receiving of goods or services

produced in another country. 𝐸4

Oil prices the cost of a crude oil barrel

𝐸5

Monetary Growth the increase in the amount of money in an

economy 𝐸6

Food prices the average price level of food prices on a global

scale. 𝐸7
Mathematical model

Problem formulation:
To implement the mathematical model aiming to assess the probability of inflation rising
in the next year, Bayes theorem will be my base of calculation in my model while taking into
account some economic indicators (EI) as conditional events. These indicators are displayed in
the table below, the first column indicates the name of the economic indicator, the second gives a
brief definition of the EI, and the last column shows the acronym of the EI that will be used in
my mathematical model.

The variables of my models are:

Let I represent the event that inflation increases in the next year.
Let 𝐸1 , 𝐸2 ,𝐸3 ,𝐸4 , 𝐸5 , 𝐸6 , 𝐸7 represent the events that CPI, PPI, GDP, International imports, Oil
prices, Monetary growth and Food prices will increase respectively

Bayes’s theorem states that:

7
7 𝑃( ⋂ 𝐸𝑖 |𝐼)· 𝑃(𝐼)
𝑃(𝐼| ⋂ 𝐸𝑖 ) = 𝑖=1
7 (1)
𝑖=1 𝑃( ⋂ 𝐸𝑖 )
𝑖=1

Where:

7
-​ 𝑃(𝐼| ⋂ 𝐸𝑖 ) Is the posterior probability that inflation will increase given that the 7
𝑖=1
economic indicators increase.
7
-​ 𝑃( ⋂ 𝐸𝑖 |𝐼) Is the probability that the 7 economic indicators will increase given that
𝑖=1
inflation increases.
-​ 𝑃(𝐼) is the prior probability that inflation increased based on historical data over 5 years.
It can be calculated using the total law probability as shown below:

7
𝑃(𝐼) = ∑ 𝑃(𝐼 ∩ 𝐸𝑖 ) (2)
𝑖=1
 7
-​ 𝑃( ⋂ 𝐸𝑖 |𝐼) is the total probability that both CPI and PPI will increase.
𝑖=1

Indeed, you can add Economic Indicators according to their importance. But in this
subsection, I will only focus on two economic indicators which are the CPI and PPI. The reason I
chose these two economic factors is because CPI is strongly correlated with the previous EIs.
Moreover, PPI will also be used because changes in PPI cause changes in CPI. The central
banks, such as the Federal Reserve, use CPI and PPI to track inflation and if both increase, they
are more likely to raise interest rates to keep the economy stable.

​ For the sake of clarity, (Clark, 1995) explained that changes in prices of oil materials
should pass through to prices of final goods and ultimately to consumer prices. To show this
connection between CPI and PPI, Rogers proposed this simplified model:

Rogers (1998)’s model about the PPI and CPI’s relationship.

Before using these indicators it is essential to know what the consumer price index (CPI)
and producer price index (PPI) mean and how they affect major global economies:

The consumer price index (CPI) is a measure of average prices for services and/or goods
purchased by consumers. The CPI commonly determines whether or not the prices of said
goods/services have increased, decreased, or stayed stable over time, in other words, whether
these prices have deflated or inflated over time. What is considered to be CPI varies in each
country and region. The rate of CPI rise or fall affects the current environment, more specifically,
when CPI rises, the probability that inflation will increase.

The producer price index (PPI), is the measure of inflation from the perspective of
producers. similarly to the CPI, also calculates the average prices of goods, but calculates the
prices of goods sold to producers to create goods and services used by consumers, and in terms
does not include sales tax, PPI usually measures the prices of raw goods and labor services. In
parallel to CPI, PPI varies in different regions and countries but fundamentally covers the price
changes in three stages: raw materials, intermediate goods, and finished goods.
 The previous generalized mathematical model can be simplified to this more specific
model, incorporating only two economic indicators, the CPI and the PPI:

𝑃(𝐶∩𝑃|𝐼) · 𝑃(𝐼)
𝑃(𝐼|𝐶 ∩ 𝑃 ) = 𝑃(𝐶∩𝑃)
Where:

-​ 𝑃(𝐼|𝐶 ∩ 𝑃 ) Is the posterior probability that inflation will increase given that CPI and the
PPI increases.
-​ 𝑃(𝐶 ∩ 𝑃|𝐼) Is the probability that both CPI and PPI will increase given that inflation
increases.
-​ 𝑃(𝐼) is the prior probability that inflation increased based on historical data over 10
years.
-​ 𝑃(𝐶 ∩ 𝑃) is the total probability that both CPI and PPI will increase.

Empirical study:

To assess the effectiveness of my mathematical model, I used the data available at the
Federal Reserve Economic Database [1], and mospi.gov and used the Arithmetic average
𝑛
1
formula: 𝑥 = 𝑛
∑ 𝑥𝑖 where 𝑥𝑖 is the measure of CPI/PPI within the data between 1st September
𝑖=1
2014 to 1st august 2024, and n is the total measures of CPI and PPI during the same period, and
2
Σ(𝑥𝑖−𝑥)
the standard deviation that can be calculated using the formula: σ = 𝑛
to gather correct
data from 2014-2024.

Economical Indicators

Country CPI (Per year) PPI (Per year)

Average Standard Dev. Average Standard Dev.

(Per year) (Per year) (Per year) (Per year)

United States of {236.763, {1.059, 1.187, {117.300, {1.153, 0.630,

America 238.625, 243.407, 1.217, 1.840, 119.067, 120.683, 0.333, 4.062,
 249.252, 254.004, 1.446, 1.224, 125.825, 129.775, 0.550, 3.135,
257.762, 265.296, 4.640, 7.489, 134.042, 140.163, 3.904, 4.212,
285.809, 301.374, 3.018, 2.579}. 154.887, 160.672, 2.570, 0.039}.
310.974}. 164.137}.

{127.74, 129.91, {0.75, 0.72,0.78, {104.5, 101.2, {0.234, 0.167,

India 131.96, 134.01, 0.63, 0.66, 0.88, 103.8, 108.9, 0.171, 0.203,
135.98, 138.14, 0.82, 0.72, 0.89, 113.5, 118.2, 0.178, 0.153,
140.37, 142.59, 0.78}. 121.4, 127.3, 0.155, 0.193,
144.88, 147.37}. 135.2, 145.5, 0.169, 0.174,
154.8}. 0.401}.

Inflation. (rate x100)

United States of America India

{162.222, 11.862, 126.158, 213.011, {666.565, 490.697, 484.821, 332.817,

244.258, 181.221, 123.358, 469.785, 393.882, 372.950, 662.343, 513.140,
800.279, 411.633}. 669.903, 564.914}.

To indicate whether these economic indicators in both of these countries have increased, I set a
threshold on the table below as a baseline. For example, If the inflation has surpassed the value of 400, it
will be considered as an increase.

Inflation CPI PPI

USA 410 280 150

India 400 140 140

Calculation
-​ 𝑃(𝐼) :
From the presented historical data, I have calculated how often inflation has increased in the past
ten years. The probability that inflation will increase in the next year will be:
For USA :
3
𝑃(𝐼) = 10
= 0.3
For India :
7
𝑃(𝐼) = 10
= 0.7
𝑃(𝐶 ∩ 𝑃|𝐼) : to calculate the probability, we observed how often the CPI and PPI both have
increased when the inflation increased in the same years. The probability was:
For USA :
2
𝑃(𝐶 ∩ 𝑃|𝐼) = 3
= 0.66
For India :
2
𝑃(𝐶 ∩ 𝑃|𝐼) = 8
= 0.25

-​ 𝑃(𝐶 ∩ 𝑃): here, the probability calculation was based on calculating how often both the
PPI and CPI have increased in all ten past years without taking in consideration if the
inflation increased or not
For USA:
2
𝑃(𝐶 ∩ 𝑃) = 10
= 0.2
For India:
2
𝑃(𝐶 ∩ 𝑃) = 10 = 0.2

After finding the elements in the equation, to calculate 𝑃(𝐼|𝐶 ∩ 𝑃 ), I plugged in the
values I have found for both the USA and India:

Plugging in the values I’ve found for both countries:

0.66×0.3
USA: 𝑃(𝐼|𝐶 ∩ 𝑃 ) =
0.2
= 0.99
0.25 ×0.7
India: 𝑃(𝐼|𝐶 ∩ 𝑃 ) = 0.2
= 0.875

These results show that in the USA, there is a 99% chance that inflation will increase in
the next year and a 87.5% increase in India, given that CPI and PPI will increase. These values
suggest that based on historical data there is a strong correlation between the increase in inflation
and CPI and PPI in both countries. Consequently, these probabilities are quite high indicating
that the CPI and PPI are good determinants to indicate the increase in inflation. Though both
these indicators participate in the increase in inflation, external factors like the pandemic and
geopolitical events may affect the inflationary pressure.
Evaluation

Limitations.
My empirical tests were heavily based on historical data for both the USA and India, but
an issue arose when searching for data for its CPI and PPI, where data given to us only lasted up
to the year 2017, to deal with this issue, I had to use an interpolation method to complete the data
from 2017-2024, but using interpolation may not be ideal for data with large gaps in between
these years mostly when there are rapid changes, periods of unexpected events, or turning points.
One other limitation I did not take into consideration, is some external factors that might affect
inflation, like the Covid-19 pandemic, the 2016 market crash, or any extreme weather event, or
any geopolitical event happening in India. These limitations might be the reason why I received a
large probability of inflation increasing.

Improvements

Changes I could have implemented into my experiment to get more accurate results
would be applying prediction models like the ARIMA model or an Outliers Model to eliminate
outliers that might have been present in my data. Additionally, if we had incorporated more
economic indicators within my calculation, there might be a possibility that my result would’ve
been heavily skewed, changed, or decreased. Especially the GDP, which is a huge part of a
country’s economy and monetary management, as well as the unemployment percentage, which
could be a great factor in how people consume and produce, and the rate of inflation, since less
employment leads to more inflation.
Conclusion

Using Bayes' Theorem, I established there is a 99% chance that inflation will surge in the
United States in the coming year and a probability of 87.5% that it will do the same in India
during this period, assuming the advancement of both the Consumer Price Index and the
Producer Price Index. These results indicated a strong association of trends in inflation with
these two key economic indicators and demonstrate their usefulness in predicting inflation in the
future.

That said, there were some challenges with the analysis. For example, interpolation was
used to estimate the data from 2017-2024, which cannot capture sudden changes or other
unforeseen events. Other factors that could influence inflation and were not taken into account
include global pandemics, market crashes, and geopolitical changes. Perhaps if more indicators
were used, such as GDP or unemployment rates, a clearer and more accurate picture could have
been seen.

​ ​ ​

*
Sources:
« CONSUMER PRICE INDEX NUMBERS ON BASE 2012=100 FOR RURAL, URBAN AND

COMBINED FOR THE MONTH OF DECEMBER 2023 ». www.mospi.gov.in, MINISTRY OF

STATISTICS AND PROGRAMME IMPLEMENTATION NATIONAL STATISTICAL OFFICE,

12 janvier 2024,F.in/sites/default/files/press_release/CPI_PR_12Mar24.pdf.

Consumer Price Index for All Urban Consumers : All Items in U.S. City Average. 12

février 2025, fred.stlouisfed.org/series/CPIAUCSL.

« CONSUMER PRICE INDEX NUMBERS ON BASE 2012=100 FOR RURAL, URBAN AND

COMBINED FOR THE MONTH OF FEBRUARY 2024 ». www.mospi.gov.in, MINISTRY OF

STATISTICS AND PROGRAMME IMPLEMENTATION NATIONAL STATISTICAL OFFICE,

12 mars 2024, mospi.gov.in/sites/default/files/press_release/CPI_PR_12Mar24.pdf.​

GeeksforGeeks. « Real Life Applications of Bayes Rsquo ; Theorem ». GeeksforGeeks, 17 May

2024, www.geeksforgeeks.org/real-life-applications-of-bayes-theorem.

Jei. Using economic indicators to create an empirical model of inflation | Journal of

Emerging Investigators. emerginginvestigators.org/articles/22-076.

Macrovar. « Understanding the Producer Price Index (PPI) : Key Insights and Trends | MacroVar

». MacroVar | MacroVar Analyses Financial And Economic Data To Help You Make The Right
Financial And Business Decisions. Sign Up Free To Access Professional Financial Advice., 13

septembre 2023, macrovar.com/india/producer-price-index.​

Retail Inflation Eases to 5.10 per Cent in January, 2024. New Delhi, Inde,

https://www.mospi.gov.in/, 12 février 12apr. J.-C.,

static.pib.gov.in/WriteReadData/specificdocs/documents/2024/feb/doc2024212310601.pdf.​

Yilmaz, Sevgi Coskun. « Investigating Factors Influencing Inflation in the USA ». Equinox

Journal Of Economics Business And Political Studies, septembre 2023,

https://doi.org/10.48064/equinox.1339198.​