Actually there are lots of gravitational waves; we just haven’t detected them directly yet. The LIGO and VIRGO collaborations have put their heads together — over 700 authors! — and come up with the best limit yet on gravitational waves from inspiralling massive black-hole binaries.
The LIGO Scientific Collaboration, the Virgo Collaboration (722 authors)
(Submitted on 18 Feb 2011)
We present the first modeled search for gravitational waves using the complete binary black hole gravitational waveform from inspiral through the merger and ringdown for binaries with negligible component spin. We searched approximately 2 years of LIGO data taken between November 2005 and September 2007 for systems with component masses of 1-99 solar masses and total masses of 25-100 solar masses. We did not detect any plausible gravitational-wave signals but we do place upper limits on the merger rate of binary black holes as a function of the component masses in this range. We constrain the rate of mergers for binary black hole systems with component masses between 19 and 28 solar masses and negligible spin to be no more than 2.0 per Mpc^3 per Myr at 90% confidence.
Note the caveats on the analysis: for one thing, it’s looking for the inspiral phase in particular (which should be the easiest to see). More importantly, they’re looking for a specific mass range — between 25 and 100 solar masses total in the binary system, which is rather large. (But nicely positioned for LIGO’s frequency sensitivity.) And of course they’re looking for black holes, not neutron stars (which would be less massive).
The truth is, they shouldn’t have seen anything, according to our best theoretical estimates. From the conclusions:
We did not detect any plausible gravitational-wave candidates. However we estimated our search sensitivity and were able to constrain the merger rate of the targeted sources in the nearby Universe. We established to 90% conﬁdence that the merger rate of black holes with component masses between 19Msun and 28Msun is less than 2.0 Mpc-3 Myr-1. We note that this is still about an order of magnitude higher than optimistic estimates for such systems  (see also [13, 19])
So, keep looking. They’re getting closer; the next step is to upgrade to Advanced LIGO. Once that happens, a lack of detections will be more surprising than actually detecting something.