We detect a Neptune mass-ratio ( q \simeq 8 \times 10 ^ { -5 } ) planetary companion to the lens star in the extremely high-magnification ( A \sim 800 ) microlensing event OGLE-2005-BLG-169 .
If the parent is a main-sequence star , it has mass M \sim 0.5 M _ { \odot } implying a planet mass of \sim 13 M _ { \oplus } and projected separation of \sim 2.7 { AU } .
When intensely monitored over their peak , high-magnification events similar to OGLE-2005-BLG-169 have nearly complete sensitivity to Neptune mass-ratio planets with projected separations of 0.6 to 1.6 Einstein radii , corresponding to 1.6–4.3 AU in the present case .
Only two other such events were monitored well enough to detect Neptunes , and so this detection by itself suggests that Neptune mass-ratio planets are common .
Moreover , another Neptune was recently discovered at a similar distance from its parent star in a low-magnification event , which are more common but are individually much less sensitive to planets .
Combining the two detections yields 90 % upper and lower frequency limits f = 0.37 ^ { +0.30 } _ { -0.21 } over just 0.4 decades of planet-star separation .
In particular , f > 16 \% at 90 % confidence .
The parent star hosts no Jupiter-mass companions with projected separations within a factor 5 of that of the detected planet .
The lens-source relative proper motion is \mu \sim 7 – 10 { mas yr ^ { -1 } } , implying that if the lens is sufficiently bright , I \lesssim 23.8 , it will be detectable by the Hubble Space Telescope by 3 years after peak .
This would permit a more precise estimate of the lens mass and distance , and so the mass and projected separation of the planet .
Analogs of OGLE-2005-BLG-169Lb orbiting nearby stars would be difficult to detect by other methods of planet detection , including radial velocities , transits , or astrometry .