Latest price
—
—
12-month forecast
—
—
Probability of decline (12m)
—
Monte Carlo simulation, 10,000 paths
95% 1-year Value at Risk
—
—
Champion model accuracy
—
M6 ElasticNet, 24-month test window
Share price: history & 12-month forecast
Monthly adjusted close, with 12-month moving average and the M3 ETS forecast fan (80% / 95% intervals)
12-month scenario & risk
Monte Carlo simulation, block bootstrap of historical returns, seed 42
What moves Maybank's price
Correlation of monthly returns with sector & macro drivers
Recommendations to Management
How to read this dashboard
The honest finding
On a true one-step-ahead basis, no model in this project statistically beats a naive random-walk forecast (Diebold-Mariano p = 0.30), and no directional classifier meaningfully beats the majority-class baseline either. Maybank behaves close to an efficiently-priced, liquid stock at short horizons. Treat every number above as a risk-managed planning input, not a guaranteed outcome.
Two models, two jobs
M6 ElasticNet is the strongest explanatory model, using real historical macro and sector data, for understanding what currently drives the price. M3 ETS, validated by a 36-month walk-forward backtest, produces the genuine forward-looking 12-month forecast shown above, because it extrapolates safely where the ML models were found to become unstable.
Refresh cadence
This dashboard is generated from a reproducible Python pipeline (data pull through dashboard feed) and should be refreshed monthly as new price and macro data arrive, both to update the forecast and to re-validate that no model has started to reliably beat naive.