The pricing anomalies along with the cyclical pattern concerning unseasoned equity issues, otherwise known as the New Issue Puzzle (NIP), has been a major question mark among scholars for the past four decades. Despite numerous attempts to solve the three different fragments that make up this conundrum, there always seem to be parts that remain unresolved. An anomaly that raises controversy in particular is the IPOs long-run underperformance. Although this abnormality has been widely acknowledged there are claims that the underperformance might be due to a "bad model problem" causing the underperformance to appear more drastic than it really is. Considering these claims, this paper applies a method that has so far found relatively little application in this field, known as Propensity Score Matching (PSM). Most of the existing research on this topic has applied the same methods, providing rather similar results, but more importantly, suffering from the same limitations. This paper discusses how the PSM method can be used in order to provide more appropriate control firms by adding several new variables to the conventional matching approach based only on industry and firm size. The PSM allows us to overcome the so-called curse of dimensionality, and by doing so provides more appropriate matches when comparing the returns of relevant firms. This econometric approach is then applied to the same initial dataset used in one of the seminal studies of the NIP by Jay R. Ritter (1991, Journal of Finance), where he provides evidence in favour of the initial IPOs underpricing, the "hot issue", as well as the long-run underperformance. In this study we therefore investigate the US IPO market in the 10-year period between 1975 and 1984, which was a popular decade for investigation by several studies regarding the IPO pricing anomalies. Using the same data constitutes a good source of comparison for the results found in this research and allows us to be more confident about the origin of possible discrepancies of our results from other studies. Our research is based on the very same initial dataset as Ritter (1991), where a total of 1,526 IPOs were observed in the 36 month period following the equity issue. However, as a result of missing data on specific variables, only 797 of the IPOs were selected for the PSM approach. We further extended the dataset to include an additional period of 24 monthly returns for robustness purposes, allowing the observation of post-issue performance over a total of five years. Our evidence provides similar results to Ritter (1991), and the vast literature that followed, regarding the short-run underpricing, where also an equivalent "hot issue" pattern emerges. However our results for the long-run underperformance lead to a drastically different conclusion compared to that of Ritter (1991): the PSM method points to significantly lower underperformance, which is reduced by approximately 25%. In fact, compared to Ritter (1991), who reports a staggering percentage underperformance of -27.39% in his complete sample, we report a modest underperformance of -2.2% using a comparable sample. The results seen in this study corroborate the arguments of Cheng (2003) where she provides convincing evidence of a link between a "bad" matching approach, and a high degree of underperformance. Our results have implications for the existing theoretical explanations, like the fads theory, linking the initial underpricing and the long-run underperformance, and purporting the view that overoptimism and fads are a suitable alibi for the long-run underperformance. Taking the considerable underperformance in Ritter (1991) into account, such a connection would seem very plausible, however, the results of this study shows that the underperformance is practically unnoticeable, hence the relationship between the underperformance and the initial underpricing fades, casting doubts on the theory. The remaining underperformance after the PSM approach could still be due to fads. Although, seeing how the underperformance is rather negligible, this makes the conventional explanation rather unconvincing, and at best it will show that fads have a very limited impact. Finally, prior research has established that the underperformance narrows and eventually disappears by the fifth year. Interestingly, the results of this paper suggest a receding trend regarding the underperformance, which is fading away but not totally disappearing five years after the IPO. [ABSTRACT FROM AUTHOR]