XML is a key standard for manipulating data on the Internet. However, querying large volume of XML data represents a bottleneck for several data intensive applications. Many modern applications require processing of massive streams of XML data, creating difficult technical challenges. Among these is the optimization of XPath query processing and accurate cost estimation for these queries when processed on a massive steam of XML data. In this paper, we present a novel performance prediction model which a priori estimates the cost of any Forward XPath structural in terms of space used and time spent. The model consists of (1) a lazy stream-querying algorithm LQ (2) a mathematical performance model (linear regression functions), and (3) a new selectivity estimation technique. Extensive experiments on both real and synthetic data sets show that our model achieves accuracy better than existing approaches. The resulting prototype supports the a priori design of efficient queries, as well as automatic query optimizations.