Uncertainty is everywhere in economics, but it is surprisingly difficult to measure. We can observe inflation, interest rates, GDP, imports, and credit, but how do we observe uncertainty itself?
One possibility is to look at what people search for online.
A recent Nepal Rastra Bank working paper, Macroeconomic Effects of Economic Policy Uncertainty: Evidence from Nepal, by me and co-authors, Birendra Bahadur Budha and Swostik Nepal, explores this idea. Using Google Trends data from January 2011 to April 2026. The paper develops a monthly measure of economic policy uncertainty for Nepal and examines what happens when uncertainty rises.
1. The problem: uncertainty is difficult to measure
Uncertainty is not directly observable. Traditional measures often rely on newspaper coverage, surveys, financial-market indicators, or statistical models.
For Nepal, high-frequency measures of uncertainty are particularly limited.
This raises a simple question:
Can people's search behavior provide a signal of how uncertain they are?
2. The idea: people search when they are uncertain
When an important economic event occurs, or is about to occur, people often look for information.
They may search for the budget, monetary policy, interest rates, foreign exchange, taxation, banking regulations, or other economic issues.
Google Trends allows us to observe changes in the intensity of these searches.
The intuition is straightforward: greater concern about economic policy may generate greater information-seeking, which can show up in search behavior.
3. Building Nepal's Google Uncertainty Index
We use 56 keywords related to fiscal, monetary, and trade policy to construct a monthly Economic Policy Uncertainty Index.
Google Trends data are normalized using “Nepal Rastra Bank” as the benchmark search term. The raw series is then adjusted for seasonality and long-term trends.
We also construct broader indices using 95 and 201 keywords to test whether the results depend on the particular set of search terms.
4. What does the index tell us?
The index rises around several major economic and political episodes.
The broader uncertainty index, for example, captures spikes around the 2013 election, the 2015 earthquake, subsequent elections, and the COVID-19 pandemic.
The economic policy uncertainty index also responds to episodes such as budget and monetary policy announcements and foreign-exchange reserve pressures.
This provides some reassurance that the index is capturing periods when economic agents were paying greater attention to uncertainty.
5. What happens after an uncertainty shock?
The paper then asks the more important question:
Does uncertainty actually affect the economy?
Using local projections, we find that a one-standard-deviation increase in economic policy uncertainty is followed by:
- around a 2% decline in GDP at the peak;
- around a 15% decline in real imports;
- around a 5% decline in private-sector credit; and
- a substantial decline in the NEPSE index.
The response of CPI, however, is not statistically significant.
The effects are also temporary, with economic activity gradually recovering as the uncertainty shock fades.
6. Why does the stock market react before GDP?
One of the most interesting findings is the difference in timing.
The NEPSE responds almost immediately, while the effect on GDP emerges with a lag of roughly two quarters.
This is intuitive. Financial markets can incorporate new information almost instantly. Investors can change their portfolios today. But businesses need time to postpone investment, adjust production, or change borrowing decisions. Households and banks similarly adjust their behavior gradually.
Financial markets therefore appear to be an early transmission channel through which uncertainty affects the broader economy.
7. What does this mean for policymakers?
The potential policy value of the index lies in its frequency.
GDP tells us what happened in the past. Google searches can provide information about what people are paying attention to right now.
A sudden increase in searches related to monetary policy, foreign exchange, taxation, banking, or other economic issues could therefore provide an additional signal of rising uncertainty.
This could make Google Trends useful as a complementary indicator for economic monitoring and nowcasting.
It is not a replacement for official statistics. Rather, it can provide an additional, timely source of information.
A new window into Nepal's economy
The interesting contribution of this research is not simply that Google searches can be turned into an index.
It is that search behavior may provide an early window into how economic agents are responding to uncertainty.
The evidence suggests that when uncertainty rises, financial markets react first, followed by declines in imports, credit, and economic activity.
In an environment where timely information matters for policy decisions, even a noisy real-time signal can be valuable.
Sometimes, before an economic shock appears in the statistics, it may already be visible in what people are searching for.
We are thankful for coverage by leading media outlets, including The Kantipur Daily, The Kathmandu Post, Onlinekhabar, Clickmandu, and many others.
https://ekantipur.com/business/2026/08/04/policy-uncertainty-negatively-impacts-the-economy-nepse-and-import-trade-are-most-affected-26-27.html
https://kathmandupost.com/money/2026/08/02/policy-uncertainty-hit-economy-as-stocks-and-imports-suffer-most
https://www.myrepublica.nagariknetwork.com/news/economic-and-policy-uncertainty-hit-hard-nepals-macroeconomic-progress-nrb-67-87.html
https://www.onlinekhabar.com/2026/08/1987662/impact-of-special-events-on-stock-market-first-study
https://clickmandu.com/2026/07/478687.html
https://www.capitalnepal.com/detail/83221
https://www.bikashnews.com/story/573270/
https://www.bizshala.com/article/30916
https://www.corporatenepal.com/story/284467
https://www.beemapost.com/2026/07/224920/
Post a Comment