What is HALO? 

HALO is dedicated to making sure that no one runs out of money in retirement due to longevity (outliving assets), health, and eldercare costs- one cause of bankruptcy in retirement. 
HALO’s Scientific Data Evaluation Process
We use clinically proven peer-reviewed longitudinal studies as a prescriptive proxy since individual statistical datasets on any given person do not exist. Our process evaluates a vast amount of data which is then carefully vetted based on criteria relating to source & data quality.
The evaluation criteria includes:
  • Relevance of study (reputation of resource, when it was done, where it was done, etc.
  • Authorities/experts behind the research
  • Publication Impact factor
  • Study design (looking for flaws in the approach)
  • Clinical Significance
  • Correlations between different risk factors and how they control for those factors and how they treated each factor separately (for example, BMI by age, gender, ethnicity)
  • Hierarchy of evidence (focusing on secondary reviews published in peer-reviewed, high-impact journals and high-quality randomized controlled trials with definitive results).
If there are multiple studies that pass this vetting process, we select the data with the
largest sample size and highest publication reputation.
HALO’s Data Sources (Sampling)
The end result of the evaluation process is data from 90 different highly vetted sources, including:
  • CDC– Centers for Disease Control
  • CMS–Centers for Medicare & Medicaid Services
  • NIH– National Institute of Health
  • SSA–Social Security Administration
  • JAMA–Journal of the American Medical Association
  • SEER (Surveillance, Epidemiology, and End Results) Cancer Research
  • Kaiser Family Foundation
Support for the Outputs & Algorithm
The HALO process uses carefully screened health and financial data in a logical
fashion, applied to an individual’s self-reported health and lifestyle circumstances. The
HALO process is not a guaranteed forecast of future events.
HALO’s proprietary algorithm weighs and combines these factors to create projections.
We strongly believe these outputs and projections are much more directionally accurate
and useful than general population-based averages /actuarial tables. As part of our
process, we work with outside third parties to audit and ensure that the technical process is
We proudly stand behind the algorithm’s construction process and in the quality of the
scientific data used as inputs.
Assembling the Pieces
The Pieces: Datasets
  • Powered by 100 million data points from over 90 curated data sources
  • Carefully vetted based on criteria relating to source & data quality
  • New data and sources are tested and evaluated on an ongoing basis
The Super Glue: The Statistics
  • Established & trusted statistical approaches (e.g., Bayes’ rule)
  • Third-party audit & verification
  • Genetic Age is the basis for projecting longevity & years of disability

Methodology Steps

Step 1:
Estimate HALO’s Proprietary “Genetic Age”
Step 2:
Using Genetic Age and lifestyle factors to project life
expectancy and probability of living to future ages
Step 3:
Using Genetic Age to project total years of disability
Step 4:
Use lifespan and disability projections to create
financial projections. Together these steps form a detailed, personalized
annual assessment of the future of health & wealth.
How Ethnicity Factors into Adjusted Life Expectancy
One’s ancestry/ethnicity plays a big role in determining our expected lifespan, and this adjustment is
similar to the BMI adjustment in that it is age dependent.
For example, the difference in expected lifespans for black and white males at age 25 (- 3.6 years) is very different from the expected lifespans for blacks and whites who have already reached the age of 65 (-1.6 years). We incorporate age-dependent differences in life expectancy for whites, blacks, Hispanics, Indians, and Asian ancestry.
To calculate the ethnicity adjustment, first round the real (biological) age to the nearest 5th year (
e.g. 33 years rounds to 35 years). Then use the user’s age, ethnicity, and gender and the chart below to find the adjustment. People of European ancestry (“white people”) have zero adjustment since they are the reference population.
For example, for a 28-year-old man who is 50% European, 25% African, 25% Asian, 0% Hispanic and 0% Other, his Ancestry Adjustment would be as follows:
Ancestry Adjustment = (0.50)*(0) +(0.25)*( - 3.4) +(0.25)*(7.8) +(0)*(2.3)+ +(0)*(0) = 1.1 years
Percentage_Asian*Asian_Adjustment + Percentage_Hispanic*Hispanic_Adjustment +



Please sign in to leave a comment.

Was this article helpful?

0 out of 0 found this helpful
Powered by Zendesk