Jens Morten Hansen (JMH) and co-authors have recently published a study where they use sine-regression to fit 5 oscillations plus a linear trend to a 160-year sea level record from waters near Denmark (a stack of local GIA corrected tide gauge records). They then observe that the sines-plus-trend model correlates highly with the record it was fitted to, after applying a 19-year moving average. This should be unsurprising as the procedure guarantees high correlation, regardless of input data (see figure). This result is clearly not significant in any meaningful sense of the word.