Do negative-cashflow properties imply greater appreciation?
More or less yes, it depends on what creates the negative cash flow. Generally speaking the long run relationship between rents and house prices can be expressed as follows;
$$R_t = P_t[(i_t + τ p_t )(1 − τ y_t ) + δ_t + λ_t − E_tG_{t+1}], \tag{*}$$
where $i_t$ is the real interest rate, $τ p_t$ is the property tax rate, $τ y_t$ is the marginal
income tax rate, $δ_t$ is the combined maintenance and depreciation rate, $λ_t$ is the risk premium associated with housing, and $E_tG_{t+1}$ is expected capital gains (Gallin 2004).
The cashflow that someone who owns and rents a house would get would be in essence;
$$R_t - P_t[(i_t + τ p_t )(1 − τ y_t ) - m_t] =CF(P)$$
where $m_t$ is portion of $\delta$ spent on maintenance ($d_t$ is portion that is depreciation). Next the we can rearrange * and solve for capital gains to get;
$$-\frac{R_t - P_t[(i_t + τ p_t )(1 − τ y_t ) - m_t] }{P_t} -d_t - \lambda = E_tG_{t+1}, \tag{**}$$
Or;
$$-\frac{CF(P_t) }{P_t} -d_t - \lambda = E_tG_{t+1}$$
The equation above shows, that if the cashflow is higher then the expected capital gains from the houses are lower. However, there is a caveat. The cash flow depends negatively on $P$ but $P$ is both in numerator and denominator. Hence, if the negative cash flow is due to fall in rents, increase in interest rate, taxes or maintenance yes, if due to increase in the house price the effect depends on whether the change in numerator dominates the change in denominator.
Can it be implied by the efficient market hypothesis (EMH) and the fact that these properties otherwise yield a relatively low ROI that this higher rate of appreciation will likely continue?
I do not see how that is implied by EMH. First, efficient market hypothesis is typically a hypothesis that is applied to stock markets. Theoretically you could apply it to other asset prices I suppose but to my best knowledge it is not applied to housing market. However, suppose we were to apply it housing market.
Efficient market hypothesis just says that asset prices are unpredictable because all avaiable information is already 'priced in' the price of asset. That is not the same as saying the price of asset does not depend on some fundamentals. It just means that all knowledge about the fundamentals is already reflected in the price so any price change is due to some new unexpected information. Inasmuch I can't see how this would violate EMH if applied to house prices. EMH doesn't claim there won't be trend in the aggregate prices. It just claims the individual prices will follow random walk around a trend given by fundamentals.
Also, your question makes it seems like Cambridge house prices always increase by 0.5% faster, but your own data show that is clearly not the case. Your own data you linked show that the prices in Cambridge sometimes increase faster sometimes slower than in rest of the England. In fact I am not even sure where yo got the 0.5% figure. The site you used as a source does not mention that.
He also states that he's almost 100% sure that the housing market would drop, which also seems to violate the EMH.
Again not sure how this would violate EMH. I am almost 100% sure stock market will be increasing on average over next 50 years. Neither Sal's or mine obesrvation violates EMH. EMH says the price of individual asset is unpredictable, it doesn't say you cannot predict long term trends on stock market etc. You just can't do better than market on average.
