top of page

Gold Price Short-Term Forecast

Writer: Tian Khean NgTian Khean Ng

The Gold price has been rising lately due to greater uncertainty in the global economic macro environment. See chart above.

 Among the many factors currently affecting its price are:

1. Geopolitical Tensions and Safe-Haven Demand

  • Ongoing Conflicts: Escalations in the Russia-Ukraine war and the Israel-Hamas conflict heightened global uncertainty, driving investors toward gold as a traditional safe-haven asset.

  • Broader Instability: Fears of regional spillovers in the Middle East and concerns over Taiwan-China relations further fueled risk aversion.

2. Central Bank Purchases

  • Diversification Strategy: Central banks, particularly in China, Turkey, and India, aggressively increased gold reserves to reduce reliance on the US dollar and hedge against geopolitical and economic risks.

  • Record Buying: The World Gold Council reported sustained net purchases by central banks, reinforcing long-term demand.

3. US Dollar Weakness

  • Fed Policy Shifts: Anticipation of the Federal Reserve pausing rate hikes and potential cuts in 2024 weakened the US dollar (DXY Index trended downward), making gold cheaper for foreign buyers and boosting demand.

4. Inflation and Interest Rate Dynamics

  • Hedging Against Inflation: Persistent inflation in key economies kept gold attractive as a store of value, despite higher nominal interest rates.

  • Real Interest Rates: Low or negative real rates (adjusted for inflation) reduced the opportunity cost of holding non-yielding gold.

5. Recession Fears and Economic Uncertainty

  • Yield Curve Inversion: Continued inversion of the 2/10-year Treasury curve signaled recession risks, prompting preemptive shifts into safe assets like gold.

  • Banking Sector Stress: Lingering concerns from 2023 regional banking crises in the US eroded confidence in financial markets.

6. Investment and Speculative Demand

  • ETF Inflows: Renewed interest in gold-backed ETFs after periods of outflows indicated renewed retail and institutional demand.

  • Cryptocurrency Volatility: Bitcoin's fluctuations led some investors to revert to gold for stability.

7. Supply Constraints

  • Mining Challenges: Production bottlenecks due to labor disputes, energy costs, and environmental regulations limited supply growth, tightening the market.

8. Cultural and Seasonal Demand

  • Festival Buying: Seasonal demand from India (Diwali, weddings) and China (Lunar New Year) provided additional price support.


Short-Term Forecast

The Table below shows that Boosted Decision Trees was the most suitable model for Gold Price Forecast lowest MAD (Mean Average Deviation)


Short-term forecast Trend


A short-term forecast using our models was made. A visualization of the short-term forecast trend is in the chart above. This forecast is autoregressive and for 1-20 steps ahead. It was based on price data from the World Gold Council.  Only 66 data points from 1/11/24 was used because most of the uncertainties are related to perceptions of the new US Administration, anticipations and reactions to its implementation policy.

Observations: In the price chart at the beginning of this post, you can see that before 20 Jan 2025, there was even a brief decline in the Gold price. And it started to develop a strong uptrend from 20 Jan after the inauguration of MAGA Man. But the forecast trend chart above shows a moderation in the rate of price increase and volatility for the next 20 days, as investors see that the MAGA Man has walked back, and scaled down his election promises re Tariffs (China, Canada, Mexico), and the US Supreme Court has put some of his intentions on hold e.g no citizenship by birthright, canceling of US Aid etc.

Nevertheless, there is still a lot of uncertainty. The original Probability Distribution Function (PDF) of the Gold price was really irregular and the high degree of skewness , and kurtosis made it impossible for statistical forecasting. See below:


But after determining the best model was using Boosted Decision Trees, ARIMA, and then Monte Carlo Simulation, we were able  to fit the output data into an unbounded MetaLog PDF that facilitates Forecasting. In the Metalog PDF below you can see that the right-hand side of the PDF i.e. the higher Gold Prices has a longer tail denoting a higher standard deviation from the Mean which means a higher risk level for forecasting.



Support, Resistance and Predictability.

The Table below shows that at the 5% quantile the Support price is 2756, and at the 95% quantile, the Resistance is 2834. Since the Support price is more predictable as the PDF shows, the Support is thus a Buy price while price levels after the Mean price of 2795 are opportunities for take profit according to your risk profile.  



 
 
 

Comentarios


bottom of page