In recent years, the Ghanaian Cedi (GHS) has experienced notable appreciation against major global currencies,

coinciding with fluctuating domestic fuel prices. Understanding the sustainability of this appreciation is crucial

for policymakers, investors, and economic planners. This study adopts an innovative interdisciplinary approach

interpretations from Ancient Geomantic traditions particularly those relating to elemental balance and directional

energy flows to provide a qualitative framework for interpreting economic cycles and currency stability. The

integration of these two paradigms allows for a richer, multi-dimensional understanding of economic

phenomena. Our findings suggest that while Machine Learning models offer robust predictive capabilities,

Geomantic insights contribute contextual depth, potentially revealing underlying patterns not captured through

conventional quantitative methods alone. This study contributes to the growing discourse on blending traditional

knowledge systems with modern computational tools in financial and economic analysis.

economy, LSTM, Random Forest, Prophet.

Service (2023) says that fuel prices have gone up by more than 40% in the last 18 months. This is mostly because

of changes in the global oil market and taxes in Ghana. These are conditions that usually lead to lower investor

confidence and capital outflows (Ofori-Abebrese et al., 2021).

Policymakers, financial analysts, and investors need to know what is going on behind this strange rise in value.

Is it a short-term correction or a sign of deeper structural resilience? (Gockel et al., 2022). The link between fuel

prices and exchange rates is complicated and affected by a number of macroeconomic factors, such as inflation,

interest rates, the balance of payments, fiscal policy, and investor sentiment (Ghosh et al., 2016; Bahmani-

Oskooee & Saha, 2019). Traditional econometric models like Vector Autoregression (VAR) and Autoregressive

Distributed Lag (ARDL) often don't work well in emerging markets like Ghana, which have high volatility and

other methods on financial time series data that is high-dimensional, noisy, and non-linear (Hastie et al., 2009;

Mullainathan & Spiess, 2017). We use Long Short-Term Memory (LSTM) networks, which are a type of

recurrent neural network (RNN) that are good at finding long-term dependencies in sequential data (Hochreiter

& Schmidhuber, 1997; Greff et al., 2015). We also use Random Forest Regression, which is known for being

robust against overfitting and able to handle features that are different from each other (Breiman, 2001;

Wihartiko et al., 2017), and Prophet, a model that Facebook made for predicting time series with strong seasonal

patterns (Taylor & Letham, 2018). The Bank of Ghana (2023) and the World Bank (2023) provided the historical

data that these models were trained on. It includes exchange rates, fuel prices, inflation, interest rates, and trade

balances from 2015 to 2023.

Warnock, 2005). These kinds of epistemologies, which are often ignored in mainstream economic discussions,

provide a qualitative, cyclical, and holistic way to understand economic events (Nkrumah, 1970; Wiredu, 1980).

We want to create a dual-lens analytical framework that combines computational forecasting with ancestral

wisdom by comparing geomantic readings from traditional practitioners in the Ashanti and Ewe regions with the

outputs of ML models (Odame, 2007; Nyamekye, 2021). For example, certain geomantic patterns linked to

"stability" (like Oyeku in Ifá) or "flux" (like Ogbe) are linked to times when the currency goes up or down,

respectively, to look for symbolic connections with economic cycles (Abimbola, 2005; Adewale, 2012). This

method is similar to recent calls for epistemic pluralism in economics and development studies (Kuada, 2010;

Mkandawire, 2005; Nnadozie, 2003), which stress how important it is to include local knowledge in scientific

digital technologies to give new perspectives (Dei, 2014; Zuberi, 2004).

The main question this study tries to answer is whether the current value of the Ghana Cedi can last even though

fuel prices keep going up and the economy as a whole is unstable. We look at both statistical probabilities and

symbolic signs of economic resilience by comparing ML forecasts with geomantic readings. According to early

results, ML models predict continued short-term appreciation due to more foreign investment and better trade

balances (Bank of Ghana, 2023). However, geomantic readings show underlying energetic imbalances that could

lead to future volatility, which is in line with the idea of karmic economic cycles found in some African

cosmologies (Mbiti, 1990; Tempels, 1959).

In the past few weeks, the value of the Ghanaian currency, the Ghana Cedi (GHȻ), has changed a lot (Bank of

Ghana, 2023). Over the past few months, the Cedi has unexpectedly gained value against major global currencies

like the US Dollar (USD) and the Euro (EUR). This is a change from a long-term trend of depreciation that

started in 2020 (Agyapong et al., 2021; Quartey & Turkson, 2022). This rise in value is especially surprising

because fuel prices at home are also going up, which has historically caused currency depreciation because it

puts inflationary pressure on the economy (Ackah & Asomani, 2019; Amoani et al., 2020). The Ghana Statistical

Service (2023) says that fuel prices have gone up by more than 40% in the last 18 months. This is mostly because

of changes in the global oil market and taxes in Ghana. These are conditions that usually lead to lower investor

confidence and capital outflows (Ofori-Abebrese et al., 2021).

Policymakers, financial analysts, and investors need to know what is going on behind this strange rise in value.

interactions between these variables (Adenutsi, 2011; Mensah & Tweneboah, 2020). Diks et al. (2011) say that

linear models may not take into account regime shifts and feedback loops that happen a lot during times of

economic instability.

To get around these problems, this study uses Machine Learning (ML) methods to model and predict how

changes in the value of the Cedi will affect fuel prices. ML algorithms have been shown to work better than

other methods on financial time series data that is high-dimensional, noisy, and non-linear (Hastie et al., 2009;

Mullainathan & Spiess, 2017). We use Long Short-Term Memory (LSTM) networks, which are a type of

recurrent neural network (RNN) that are good at finding long-term dependencies in sequential data (Hochreiter

& Schmidhuber, 1997; Greff et al., 2015). We also use Random Forest Regression, which is known for being

This paper goes beyond traditional quantitative analysis by using ideas from Ancient Geomancy, a traditional

way of interpreting space and symbols that has deep roots in African indigenous knowledge systems (Mbiti,

1990; Asante, 2003). Ifá (Yoruba), Sikidy (Malagasy), or Dakpɛ (Ewe) are some of the names for geomancy in

different African cultures. It is based on the idea that natural and cosmic energies affect people's lives and the

rhythms of society (Bascom, 1969; Ellis, 1890; Gyekye, 1996). To understand cycles of good luck, stability, and

change, these systems often use symbolic arrangements, like the 16 main figures in Ifá divination (Idowu, 1962;

Warnock, 2005). These kinds of epistemologies, which are often ignored in mainstream economic discussions,

provide a qualitative, cyclical, and holistic way to understand economic events (Nkrumah, 1970; Wiredu, 1980).

We want to create a dual-lens analytical framework that combines computational forecasting with ancestral

wisdom by comparing geomantic readings from traditional practitioners in the Ashanti and Ewe regions with the

fuel prices keep going up and the economy as a whole is unstable. We look at both statistical probabilities and

symbolic signs of economic resilience by comparing ML forecasts with geomantic readings. According to early

results, ML models predict continued short-term appreciation due to more foreign investment and better trade

balances (Bank of Ghana, 2023). However, geomantic readings show underlying energetic imbalances that could

lead to future volatility, which is in line with the idea of karmic economic cycles found in some African

cosmologies (Mbiti, 1990; Tempels, 1959).

when looking at financial events. Unwin (1998) and later Adebanwi (2012) both said that the future of African

Moreover, while ML-based forecasting methods are gaining traction in financial modeling, they often overlook

by proposing a hybrid analytical framework that integrates quantitative machine learning predictions with