Business resilience has gained prominence, in academia and practice, vis-à-vis the heightened challenges recently faced by organizations, e.g. financial crisis. Developing resilience by thriving or bouncing back from crises yields sound business health in the future.However extant scholarly discussion on predictive modelling of economic resilience is rather limited, while business health studies are mainly limited to bankruptcy failure predictions. These studies mostly utilize financial snapshots (based on only few years data) to construct the predictive models hence are static in nature (Balcaen and Ooghe 2006). Several assumptions underpin these static models, e.g. considering failure as a steady process devoid of organizational history (Appiah et al. 2015, du Jardin and Séverin 2011). Even though, few recent studies (cf. du Jardin and Séverin (2011), Chen et al. (2013) etc.) have designed a “trajectory of corporate collapse” to forecast the changes in firms’ financial health, using various ‘expert systems’ like self-organizing maps (SOM) based upon unsupervised neural network approach, these studies still interpret the findings largely for predicting bankruptcy (a ‘state’) rather than drawing inference on the economic growth or recovery patterns (a ‘trajectory’) of organizations – a key to generate resilience. Neither these studies utilize longitudinal financial data (spanning over many years) to capture the dynamics of corporate history required to build resilience of organizations in reality.In this context, our paper proposes developing a predictive econometric model of business resilience by using ‘expert’ SVM method. The expanded predictor based on financial ratios highlighted by Altman (1968)’s Z-score also takes into consideration the corporate dynamics (first and second derivatives). Historical financial data is gathered from 198 firms representing 26 Dow Jones industrial sectors, and starting from 1960s.Our prediction model achieved comparatively high predictive accuracy of ---- (for a forecasting horizon of ----- years) and is comparable to similar studies. However, the main contribution of the paper is in proposing four archetypical patterns in business health trajectories, derived from the historical hind-sight, defined by tendency-dynamics combinations and is essential to characterize business resilience as follows:

Business Health (at T = t+1) = Business Health (T = 0 to t) + Resilience function

These four typical situations range from the most pessimistic case (tendency = Down, dynamics = Down) to the most promising (Up-Up). The four archetypes can be used to explain four resilience functions, viz. (i): up-up as sustainable resilience, (ii) up-down as short-term resilience, till t = T, (iii) down-up as resilience in near-future, at t = T, and (iv) down-down as lack of resilience.