The regression was estimated for each sector separately by applying the logit model:
where denotes the speculative bubble dates for sector i, which is equal to one if there is a bubble, and to zero otherwise; FPI is foreign portfolio investment; CDS is Sovereign Credit Default Swaps for Turkey; EX is foreign exchange rate (dollar/lira); VIX is CBOE Volatility Index; ∆ is the difference operator; are vectors of parameters to be estimated, and is an i.i.d. error term.
Hypothesis 1: The possibility for the existence of speculative bubbles in assets increase with higher Sovereign Credit Default Swaps (CDS), holding other variables constant.
Hypothesis 2: An increase in foreign exchange rate leads to speculative bubbles, holding all other independent variables constant.
Hypothesis 3: Foreign portfolio investment is associated with an increased probability of speculative bubbles, holding other variables constant.
Hypothesis 4: A lower CBOE Volatility Index (VIX) reduces thechance of the speculative bubble in stock market, holding the effect of the other variables fixed.
I need the interpretation of the following table (LOGIT (MARGINAL EFFECT) model results). How independent variables impact the probability of speculative bubble for each sector? I need some details like what are the reasons behind that conclusion?
I mean one unit or one per cent change in VIX index cause ….% change in the probability of bubble in Y sector, this is because…..(give me some logical explanation).
Table. The Estimation Results of Marjinal Effects
|BIST Real Est.||-0.158**||-0.207**||-0.268||0.105**|
|BIST Food Bev.||-0.008||-0.488***||0.070||0.149*|
|BIST Wood Paper||0.033||0.062||0.187||-0.116*|
|BIST Metal Prod.||-0.038||-0.163||0.027||-0.151|
|BIST Basic Metal||0.178||0.042||0.621||-0.405***|
|BIST Textile Le.||-0.015||-0.530***||0.252||0.020|
|BIST Inf. Tech.||0.169||-0.161*||0.126||0.121|
Notes: ***, **, and * denote significance levels at 1 per cent, 5 per cent and 10 per cent, respectively.