pooler error: query_wait_timeout in pgbouncer logs

I am getting pooler error: query_wait_timeout for different databases in the pgbouncer log 6 to 7 times a day . When the error occurs for masterdata database , it gives me a timeout in the application. The configuration is below :

: "md5"
  default_pool_size: 10
  max_client_conn: 5000
  max_db_connections: 400
  min_pool_size: 0
  pool_mode: "transaction"
  reserve_pool_size: 20
  reserve_pool_timeout: 1
  query_timeout: 0
  ignore_startup_parameters: "search_path, extra_float_digits"
  server_reset_query: "DISCARD ALL"
  server_reset_query_always: 1

As the query timeout is disabled (query_timeout:0) and the pool_mode is set to transaction, why am I still having this error? Do I need to set some min_pool_size for my masterdata database to eliminate the above error?