The education expenditure for children in Japan is characterized by a large share of private expenditure, especially expenditures for extracurricular study and activities. "Burden of educational costs for children" often shows up in surveys as among the top reasons why couples do not want to have children. The “Qualified Educational Fund Giving Trusts (QEFGT),” which started in 2013, aims to promote the transfer of the assets of elderly to young people, support families with children, and stimulate educational investment in the family. Under this system, the gift tax is exempted up to a certain amount when grandparents create monetary trusts to finance their grandchildren's education (including expenditures for cram school, or "Juku"). This system has been widely accepted, with the value of trust assets being 1.238 trillion yen in March 2018. This policy is unique from an international perspective as well in influencing the transfer of assets over three generations. In this paper, I conducted an empirical analysis of the association between this policy and education expenditure for grandchildren and their academic achievement using the Japan Household Panel Survey (KJHPS) and the Japan Child Panel Survey (JCPS), based on descriptive statistics and simple linear regression analysis. It is found that (1) people who recognize or use this system tend to have more financial assets than those who do not. (2) Parents of children who receive money from this system are more likely to spend on cram school or tutors, and their children are more likely to take junior high school entrance exams than parents and children who do not. (3) Controlling for several factors, there was no evidence that children receiving funds under this system had higher test scores than children who did not.