Asymptosis

Most of the data from the source is wrong, except for when they directly quote [and not massage] the data from the Fed’s FoF report [or other Fed sources], issued Qtly.

E.g. http://www.federalreserve.gov/pubs/bulletin/2012/PDF/scf12.pdf

‘Most noticeably, median incomes moved higher for retirees and other nonworking families. … The decline in mean income was even more widespread than the decline
in median income, with virtually all demographic groups experiencing a decline between
2007 and 2010; the decline in the mean was most pronounced in the top 10 percent of the income distribution and for higher education or wealth groups.’

Avg Net Worth change 2007-2010 — {ibid}

Lowest quintile +5.9% [plus 5.9 percent]
2nd Lowest: -9.5%

Middle quintile: -9.55%
2nd highest: -25.4%

80-89.9% decile: -11.2%
90-100% decile: -15.3% [minus 15.3%]

Using the Fed’s actual data, instead of…imputed + massaged guesstimates gives you a different answer.