When Katja Taipalus came home from school every day in the Finnish town of Jalasjaervi, she knew her working parents would not be there. Instead, her retired grandfather, who also lived in the large wooden house, played cards and other games with her. They even repaired a car.
Then, as she got older, there were fewer pastimes and more debates.
"He wanted to question things and enjoyed making me explain my views," said Taipalus, 39. "He didn't make it easy on me. We'd talk about absolutely everything. When I started studying macroeconomics, he was thrilled."
Those discussions taught her to stay focused when she became an economist and her research hit a dead end, she said. The end result: an indicator that helps detect asset-price bubbles in equity and housing markets - as much as a year in advance.
"Asset prices have been one of the main components as financial crises have built up," she said in the Bank of Finland's historic teller hall in Helsinki, its walls decorated with giant tapestries. "If we think of the tools needed to allow policymakers time to react," one of the main ones is "getting a signal as early as possible about an asset-price bubble in the making," she said.
Drawing up the indicator earned Taipalus her PhD last year at the University of Turku, after 4½ years of study alongside full-time work at the Finnish central bank.
She heads the macroprudential analysis division of the financial stability and statistics department, running a team of seven economists and experts who work on analysis and applied research.
The worst financial crisis since the Depression and the biggest economic slump since the second world war have prompted policymakers to develop such responses as so-called macroprudential tools, which gauge systemic risk among financial institutions and limit the impact on the economy.
Feed in dividend yields and stock indices, and Taipalus's indicator signals every major US stock-price bubble since 1871. Input rent indices and house prices, and it signals when increases in the cost of homes are becoming unhinged.
When the light flashes, as much as 12 months ahead of time, it is a sign to regulators that action may be needed. The warning is a numerical value produced by the data; its level depends on what trigger has been set.
Taipalus's indicator is able retroactively to detect the US subprime housing-market collapse after prices peaked in 2006, with signals from 1998 to 2001 and again from 2003 until early 2006.
The model also could have predicted the Spanish and Irish housing market booms before the crashes occurred. US house prices are now 23 per cent lower than the peak.
Taipalus's early-warning tool has a wide range of possible applications, from helping central bankers decide on the right timing of a cycle of monetary tightening to encouraging counter-cyclical capital buffers.
It also can help policymakers distinguish a bubble from a price increase that is driven by rising fundamentals. Taipalus now is working on extending the warning period her indicator provides.
Yale University Professor Peter Phillips has worked on similar research, with interest in the past couple of years from central banks in Hong Kong, the US, Singapore, Israel, Iceland and New Zealand as well as large commercial banks, he said in an e-mailed response to questions.
"Financial market stability, particularly global financial stability, is a public good like clean air, clean oceans and a healthy environment," he said.
"Financial stability is threatened by asset bubbles, excessive credit creation, uncontrolled risk-taking by banks and ballooning sovereign debt. Any econometric methods that assist in detecting these threats are helpful to ensuring financial stability."