You're assuming they ever cashed in, which is probably a stretch.
Not too much of a stretch!, even without personal preferences, If they were building a large scale project making use of CounterParty exclusively,
and they were investing in altcoins (both of which are confirmed truths) I can't see how it's a huge logic jump to imagine one of those altcoins might be counterparty- purely from a business perspective. I doubt many could come up with a more reasonable alternative coin they may have invested in. An internal move away from counterparty to a bitcoin-only market might correlate with sales of those 'non-bitcoin' altcoins judd bagley mentions.
Overstock listed its cryptocurrency holdings as valued at $233,000 on March 31st, down from $340,000 on 31st December.
IF they had purchased anytime after oct 2014 (Remember there was quite a large run up right after then- partially caused by the formal announcement of the Overstock/CounterParty announcement) , and sold anytime before march 31st when the report was issued (let's say february, right after the coindesk article announcing the partnership was over - and that they were pursuing other markets, was released)- the market data would also support a loss like they are reporting. In fact a block buy spike on october almost perfectly mirrors a block sell at that time.
procuring the balance sheet @
http://edgar.sec.gov/Archives/edgar/data/1130713/000113071315000018/ostk-20150331x10q.htm shows $300k investments in cryptocurrencies were made in the past 12 months but prior to the past 3 months of the filing (so before Dec 31st 2015, but after March 31st 2014), and that the sales were at some time from Dec 31st to March 31st 2015 but not prior, supporting such a theory further.
Gains or losses resulting from changes in the value of our cryptocurrency assets are recorded in Other income (expense), net in our consolidated statements of income. Losses on cryptocurrency holdings were $117,000 and zero during the three months ended March 31, 2015 and 2014, respectively.
