She first-differenced the non-stationary variables (Microfit 5 → Generate → d(x) ). Now, D(LAGOS_CONSUMPTION) and D(LONDON_REMITTANCES) became stationary. But she had lost the long-run relationship. For that, she needed Chapter 2. Chapter 2: The Long-Run Marriage (Cointegration) The PDF’s most dog-eared section was on Cointegration . "If two non-stationary series move together over time," it read, "their linear combination might be stationary. That is cointegration."
Aliyah smiled. "Short-term: strengthen remittance channels. Long-term: break the cointegration by building local savings instruments. The ECM shows you have three quarters to act before a remittance shock becomes a consumption crisis." Time series econometrics using Microfit 5.pdf
In Microfit 5: . She ordered: REMITTANCES → CONSUMPTION (remittances cause consumption, not vice versa). For that, she needed Chapter 2
The output appeared:
But the short run? That’s where the ghost hid. Microfit 5 made the Error Correction Model (ECM) seamless. From the same VAR output, she clicked View → Long Run Form (ECM) . That is cointegration
And that is the art of applied time series econometrics. The story is fictional but methodologically accurate to Microfit 5’s capabilities (cointegration, ECM, IRF, diagnostics). The actual PDF would contain step-by-step commands, screenshots, and empirical examples.